See below.

On Fri, Nov 29, 2013 at 3:31 AM, Behrens, Marion <[log in to unmask]> wrote:

Dear SPM experts,

I configured a flexible factorial design with four factors:

1st: subject
2nd: condition A, 2 factor steps
3rd: condition B, 2 factor steps
4th: condition C, 3 factor steps

For calculating/estimating the model I left out the subjects factor and just called for main effects A, B, C and interactions (AB), (AC), (BC).

>> You need the subjects main effect in the model.

Then I set up contrasts according to the Glaescher and Gitelman paper and it worked for the main effects and the interactions. But when I tried to test for single regressors I always got an "invalid contrast" message.

Let me give you two examples:

The regressors for my model are:
A1 A2 B1 B2 C1 C2 C3 A1B1 A1B2 A2B1 A2B2 A1C1 A1C2 A1C3 A2C1 A2C2 A2C3 B1C1 B1C2 B1C3 B2C1 B2C2 B2C3

To test for the single regressor A1 I put in:
1 0 0.5 0.5 1/3 1/3 1/3 0.5 0.5 0 0 1/3 1/3 1/3 0 0 0 0 0 0 0 0 0

I believe that you want:
1 0 0.5 0.5 1/3 1/3 1/3 0.5 0.5 0 0 1/3 1/3 1/3 0 0 0 1/6 1/6 1/6 1/6 1/6 1/6

** Note that this is not a valid statistical contrast because its a between-subject effect.

To test for the single regressor interaction B2C3 I put in:
0 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1

A1B2C3 contrast would be:
1 0 0 1 0 0 1 0 1 0 0 0 0 1 0 0 0 0 0 0 0 0 1
A2B2C3 contrast would be:
0 1 0 1 0 0 1 0 0 0 1 0 0 0 0 0 1 0 0 0 0 0 1

I believe that you want:
.5 .5 0 1 0 0 1 0 .5 0 .5 0 0 .5 0 0 .5 0 0 0 0 0 1

** Note that this is not a valid statistical contrast because its a between-subject effect.

You may want to start with 2 smaller models and work up to the larger contrast.

Both contrasts didn't work. Can anyone give me advice what I did wrong? Or are there simply to many factors in my model, because when I leave out factor 3, the contrasts for testing single regressors worked.

Thank you for any help in advance!

Marion Behrens

Universitätsklinikum Frankfurt
Germany

Email: [log in to unmask]