Dear SPM experts,
I have trouble to set up the factorial design in the GUI for my deconvolution model:
In principle, I have a 3 (single, repeat, switch) x 2 (congruent, incongruent) design. I specified for each of the different regressors (single_congruent, single_incongruent, repeat_congruent, repeat_incongruent and so on) an FIR model spanning 13 time bins.
However, my experiment was divided into 4 scanning runs where
- the first and second run only contained congruent as well as incongruent single-condition-trials (i.e.,2 regressors per run: single_con, single_incon)
while
- the third and fourth run contained congruent and incongruent repeat- AND switch-condition trials (i.e.,4 regressors per run: repeat_congruent, repeat_incongruent, switch_congruent, switch_incongruent).
I would like to run repeated measures ANOVA with time (1-13), switching condition (single, repeat, switch) and congruency condition (congruent, incongruent) but don't know exactly how to specify factor names and levels in the GUI to account for the fact that different runs have not the same number of condition regressors (even if the overall trial number per run ist identical across runs). Do I have to include an additional factor for run (or in SPM Terms "session")?
If I try to setup contrasts manually instead of using the automatic specification, I get trouble with these many time bins (to tap into the time by switching condition by congruency condition...)
I would be very happy about some help. Thank you very much!
Best, Sandra
|