Hi,
This looks fine as long as the groups were age-matched (primarily -
that they have the same average age), as this design cannot correct
for non-age-matching.
Don't forget to use the --T2 option for TFCE in randomise on TBSS-
preprocessed data.
Cheers, Steve.
On 29 Jan 2009, at 16:09, Matthew Hoptman wrote:
> Hi all,
> I just want to verify that I'm doing my correlational analyses
> correctly in
> randomise. I have two groups with 15 subjects in the control group
> and 29 in
> the patient group. I want to examine, separately for each group, the
> correlation between a variable of interest and FA, controlling for
> age.
>
> C P AgeC AgeP VarC
> VarP
> 1 0 15.13333333 0 1.8 0
> 1 0 -3.866666667 0 -2.2 0
> 1 0 -4.866666667 0 1.8 0
> 1 0 -12.86666667 0 -1.2 0
> 1 0 17.13333333 0 1.8 0
> 1 0 15.13333333 0 -2.2 0
> 1 0 14.13333333 0 0.8 0
> 1 0 -9.866666667 0 -0.2 0
> 1 0 -2.866666667 0 1.8 0
> 1 0 -6.866666667 0 0.8 0
> 1 0 10.13333333 0 -10.2 0
> 1 0 -4.866666667 0 1.8 0
> 1 0 -3.866666667 0 2.8 0
> 1 0 -8.866666667 0 0.8 0
> 1 0 -12.86666667 0 1.8 0
> 0 1 0 8.862068966 0 0.379310345
> 0 1 0 10.86206897 0 2.379310345
> 0 1 0 2.862068966 0 -2.620689655
> 0 1 0 -4.137931034 0 0.379310345
> 0 1 0 -8.137931034 0 -0.620689655
> 0 1 0 -6.137931034 0 4.379310345
> 0 1 0 -3.137931034 0 1.379310345
> 0 1 0 -2.137931034 0 -4.620689655
> 0 1 0 -8.137931034 0 0.379310345
> 0 1 0 5.862068966 0 0.379310345
> 0 1 0 -4.137931034 0 -0.620689655
> 0 1 0 -1.137931034 0 3.379310345
> 0 1 0 8.862068966 0 3.379310345
> 0 1 0 0.862068966 0 0.379310345
> 0 1 0 3.862068966 0 -5.620689655
> 0 1 0 0.862068966 0 -0.620689655
> 0 1 0 -7.137931034 0 -1.620689655
> 0 1 0 11.86206897 0 4.379310345
> 0 1 0 11.86206897 0 0.379310345
> 0 1 0 -11.13793103 0 -5.620689655
> 0 1 0 -4.137931034 0 1.379310345
> 0 1 0 8.862068966 0 1.379310345
> 0 1 0 0.862068966 0 -0.620689655
> 0 1 0 -11.13793103 0 2.379310345
> 0 1 0 7.862068966 0 3.379310345
> 0 1 0 -8.137931034 0 1.379310345
> 0 1 0 -12.13793103 0 -8.620689655
> 0 1 0 4.862068966 0 -1.620689655
> 0 1 0 1.862068966 0 1.379310345
>
>
> For contrasts, I have
>
> for VarC:
> 0 0 0 0 1 0
>
> for VarP:
> 0 0 0 0 0 1
>
> Does this look correct? My plan is to run this in randomise using
> TFCE. I
> assume if I wanted to look at the group difference in correlation, I
> 'd do:
> 0 0 0 0 1 -1
>
> Thanks,
> Matt
>
---------------------------------------------------------------------------
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director, Oxford University FMRIB Centre
FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------
|