Dear List,
when trying to enter cond/sess based T-contrasts in the batch editor I
observe puzzling things:
I've got a single subject design matrix with 12 sessions and 3
conditions, defined via .mat file read in as "Multiple conditions"
Condition 1 consists of one regressor (1 basis function, no time
derivative, no parametric modulation)
Condition 2 consists of three regressors (1 basis function, no time
derivative, two first order parametric modulators)
Condition 3 is only present in some sessions and consists of one regressor.
Entering in the batch editor a T-contrast for Session 1with Contrast
weight 1for
Condition # 2
Basis function # 1
Parametric modulation # 1
Parametric modulation order 0
yields a contrast vector [0 1 0 0 0 ...], which is what I'd expect,
since entering a parametric order of 0 refers to the basis function itself.
Entering
Condition # 2
Basis function # 1
Parametric modulation # 2
Parametric modulation order 0
yields the same contrast vector [0 1 0 0 0 ...], which seems to confirm:
order 0 contrasts the basis function itself
However, entering
Condition # 2
Basis function # 1
Parametric modulation # 2
Parametric modulation order 1
results in an "Error using spm_run_con (line 124)
Session-based contrast 3:
Column "Cond2 Mod2 Order1" does not exist." Same with Mod3 Order1 and
Mod3 Order0.
Can anyone explain, how these settings should be used?!
Thanks for any help,
Kai
|