Dear Judy,
Your design is unnecessary overparameterised and does not model subject
effects. Instead define three factors: subject (independent, equal), A,
B (dependent, equal, for both), and specify Main effect: 1, and
Interaction: [2 3]. Testing for main effect of A and B, and AxB
interaction will require F-contrasts (i.e. multidimensional contrasts)
because you have more than two levels per factor.
Best regards,
Guillaume.
On 06/12/2019 05:09, Judy Zhu wrote:
> Hello Everyone,
>
> I am analysing repeated-measures data in a 3x3 design (i.e. two factors,
> each with 3 levels). I am using a flexible-factorial model and have done
> the steps of "Specify 2nd-level" and "Estimate". When specifying the
> model, I included the 2 main effects & their interaction (hence, there
> are 15 columns: A1, A2, A3, B1, B2, B3, A1B1, A1B2, A1B3, A2B1, A2B2,
> A2B3, A3B1, A3B2, A3B3).
>
> Now I am at the "Results" step and am prompted to enter contrast
> matrices. I have searched through the forums but have not found an
> example of how to test a 3x3 design. I have constructed the following
> contrast matrices based on my understanding so far, and would like to
> check with you if they are correct:
>
> Main effect of Factor A:
>
> 1 -1 0 0 0 0 1/3 1/3 1/3 -1/3 -1/3 -1/3 0 0 0
>
> 0 1 -1 0 0 0 0 0 0 1/3 1/3 1/3 -1/3 -1/3 -1/3
>
>
> Main effect of Factor B:
>
> 0 0 0 1 -1 0 1/3 -1/3 0 1/3 -1/3 0 1/3 -1/3 0
>
> 0 0 0 0 1 -1 0 1/3 -1/3 0 1/3 -1/3 0 1/3 -1/3
>
>
>
> 3x3 Interaction (A*B):
>
> 0 0 0 0 0 0 1 -1 0 -1 1 0 0 0 0
>
> 0 0 0 0 0 0 0 1 -1 0 -1 1 0 0 0
>
> 0 0 0 0 0 0 0 0 0 1 -1 0 -1 1 0
>
> 0 0 0 0 0 0 0 0 0 0 1 -1 0 -1 1
>
>
> When there are multiple rows in a contrast matrix, does that mean
> multiple tests are being conducted to test that effect? Does SPM
> automatically take care of the multiple comparison problem in this case?
>
>
> Any advice would be greatly appreciated. Thank you very much in advance!
>
>
> Kind regards,
>
> Judy
>
> ---
>
> Judy D. Zhu
> PhD Candidate
> Department of Cognitive Science
> Macquarie University, Sydney, Australia
--
Guillaume Flandin, PhD
Wellcome Centre for Human Neuroimaging
UCL Queen Square Institute of Neurology
London WC1N 3BG
|