Hello everyone,
I'm trying to run a gPPI model but I'm stuck and I really need help and advice.
I have an almost block design task with 3 conditions of interest (and 3 more not of interest).
I preprocessed the data using the standard pipeline (except for smoothing, set to [6 6 6]).
The conditions were entered using SPM.mat and everything looks correct.
I have 2 groups of participants (total N = 49).
Denoising was performed using a [0.008 inf] bandpass filter and all conditions were included as confounds (as well as regressors from the ART process in the 1st and 2nd level covariates).
At the 1st level, I selected the standard gPPI, bivariate regression, and the 3 conditions of interest.
At the 2nd level, I selected the atlas.Amygdala_r.
In a first attempt to reproduce the previous analyses (lost due to a hardware problem), I compared the two groups on a cond1>cond2 contrast.
The results are very different from the previous analyses (first surprise) which showed significant differences between the groups on this contrast.
No significant activation is noted here.
To explore this result, I performed simple analyses such as the effect of cond1 in group1 and group2 separately.
Here, I notice huge clusters of activation (in attachment) that encompass the seed.
So, I'm wondering if
1) my pipeline is appropriate
2) it is common to see this type of activation on simple analyses/contracts in gPPI
Any help/advice will be appreciated!
Thanks,
Alex
I'm trying to run a gPPI model but I'm stuck and I really need help and advice.
I have an almost block design task with 3 conditions of interest (and 3 more not of interest).
I preprocessed the data using the standard pipeline (except for smoothing, set to [6 6 6]).
The conditions were entered using SPM.mat and everything looks correct.
I have 2 groups of participants (total N = 49).
Denoising was performed using a [0.008 inf] bandpass filter and all conditions were included as confounds (as well as regressors from the ART process in the 1st and 2nd level covariates).
At the 1st level, I selected the standard gPPI, bivariate regression, and the 3 conditions of interest.
At the 2nd level, I selected the atlas.Amygdala_r.
In a first attempt to reproduce the previous analyses (lost due to a hardware problem), I compared the two groups on a cond1>cond2 contrast.
The results are very different from the previous analyses (first surprise) which showed significant differences between the groups on this contrast.
No significant activation is noted here.
To explore this result, I performed simple analyses such as the effect of cond1 in group1 and group2 separately.
Here, I notice huge clusters of activation (in attachment) that encompass the seed.
So, I'm wondering if
1) my pipeline is appropriate
2) it is common to see this type of activation on simple analyses/contracts in gPPI
Any help/advice will be appreciated!
Thanks,
Alex