 |
 |
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] > >