Dear Michael,
You are right. The F-contrast testing for the interaction in 2x3
factorial design is:
[1 -1 0 -1 1 0
0 1 -1 0 -1 1]
All the contrasts can be given by calling:
c = spm_make_contrasts([2 3]);
Best regards,
Guillaume.
On 14/09/17 01:38, Michael L. wrote:
> Hi,
>
> I have a question about the contrast weights in flexible factorial. In Glascher and Gitelman's paper they use the example of a 2x3 within-subjects design with columns A1B1 A1B2 A1B3 A2B1 A2B2 A2B3. The interaction A x B is defined as -1 0 1 1 0 -1, but isn't this the interaction between the two levels of factor A (A1 and A2) and two levels of factor B (B1 and B3)?
> Thank you.
>
> Best regards,
> Michael
>
--
Guillaume Flandin, PhD
Wellcome Trust Centre for Neuroimaging
University College London
12 Queen Square
London WC1N 3BG
|