Hello Marc,
Your design with the 3 groups is a 3 x 2 x 2 ANOVA. And so the 3-way
interaction degrees of freedom equals (3-1)*(2-1)*(2-1) = 2. So
you'll need 2 contrasts to span this space. There are infinitely
many choices for these 2, but a natural choice is obtained by
starting out with the 2-way interaction term for the last two
factors:
[1 -1 -1 1 1 -1 -1 1 1 -1 -1 1]
and multiplying it element-wise by the two-pairwise group
comparsions:
[1 1 1 1 -1 -1 -1 -1 0 0 0 0]
&
[1 1 1 1 0 0 0 0 -1 -1 -1 -1]
yielding
[1 -1 -1 1 -1 1 1 -1 0 0 0 0]
&
[1 -1 -1 1 0 0 0 0 -1 1 1 -1].
Sincerely,
Eric
Quoting "M. van Ekeren" <[log in to unmask]>:
> Dear SPM-users,
>
> I have a little question about my study design.
>
> I want to compare 3 sub-groups on a 2-by-2 factorial design.
>
> For comparison between two groups I use a
> 1 -1 -1 1 -1 1 1 -1
> contrast to test for 3-way interaction
> But what to use for the 3-way interaction with 3 groups taken
> into account?
>
> So my question is whether it is possible to use an ANOVA for this
> comparison.
> and which contrast would be usefull??
>
> thanks in advance,
>
> Marc van Ekeren
>
|