Dear SPMers,
let's say one has a 2*3 design with 2 groups and 3 time points and one would like to model a group and time including the interaction in a single analysis. Though the standard full factorial is not appropriate to model directly the within subject effects, in how far would it be correct or wrong if one adjusts the full factorial by adding regressors modelling the within subject variance and setting dependence for the factor time to yes? The design matrix would then look quite the same as in any other software package (e.g. SPSS) allowing to model both directly:
F_1_1 F_1_2 F_1_3 F_2_1 F_2_2 F_2_3 regressor s1 regressor s2 regressor s2 and so on
1 0 0 0 0 0 1 0 0
1 0 0 0 0 0 0 1 0
1 0 0 0 0 0 0 0 1
1 0 0 0 0 0 0 0 0
1 0 0 0 0 0 0 0 0
1 0 0 0 0 0 0 0 0
0 1 0 0 0 0 1 0 0
0 1 0 0 0 0 0 1 0
0 1 0 0 0 0 0 0 1
0 1 0 0 0 0 0 0 0
0 1 0 0 0 0 0 0 0
0 1 0 0 0 0 0 0 0
0 0 1 0 0 0 1 0 0
0 0 1 0 0 0 0 1 0
0 0 1 0 0 0 0 0 1
0 0 1 0 0 0 0 0 0
0 0 1 0 0 0 0 0 0
0 0 1 0 0 0 0 0 0
0 0 0 1 0 0 0 0 0
0 0 0 1 0 0 0 0 0
0 0 0 1 0 0 0 0 0
0 0 0 1 0 0 0 0 0
0 0 0 1 0 0 0 0 0
0 0 0 1 0 0 0 0 0
0 0 0 0 1 0 0 0 0
0 0 0 0 1 0 0 0 0
0 0 0 0 1 0 0 0 0
0 0 0 0 1 0 0 0 0
0 0 0 0 1 0 0 0 0
0 0 0 0 1 0 0 0 0
0 0 0 0 0 1 0 0 0
0 0 0 0 0 1 0 0 0
0 0 0 0 0 1 0 0 0
0 0 0 0 0 1 0 0 0
0 0 0 0 0 1 0 0 0
0 0 0 0 0 1 0 0 0
Thank you very much for your feedback,
Juergen
|