Dear Dr. Alfonso Nieto-Castanon,

Thank you so much for this wonderful CONN tool. I hope I can get your help with a problem I had while using it.
I am attempting to examine connectivity before vs after language treatment using both seed to voxel and ROI to ROI analyses at the single subject level, with the goal of 1) identifying significant connectivity and 2) creating informative visualizations.
I carefully searched the NITRC forum for similar cases but didn't come up with a satisfying solution. Your guidance would be highly appreciated.

We have task-based fMRI data (event related design) with 3 conditions (treated words, untreated words, baseline), at two time points (pre and post). I.e., each subject has 6 conditions. I performed both SBC and gPPI analysis.
1. Is there a difference between SBC and gPPI outputs at first level?
2. For the seed-to-voxel analysis I used the scripts recommended in the NITCR forum: https://www.nitrc.org/forum/message.php?... to run a pre <� post contrast of the treated words for a single subject. I then used the second script for visualization of the .nii outputs in CONN
The resulting thresholded P_FDR map was quite strange and I'm not sure how to interpret it (png1).
Also, it was not sensitive to threshold changes, and the location of the peaks was quite different from those in the thresholded beta map for the same contrast (png2).

They were also different from the peaks at the first level effect size map of the same contrast in the CONN GUI. I think the p values for each z extracted in the script are not correct (wither 0 or .5)
3. Is there a way to do ROI-to-ROI analysis at the single subject level?
In order to use a simple ZPF to compute the difference between correlations of two conditions (e.g., treated pre vs treated post), we need to know the N in order to compute the standard error (se). I assume N would be the number of trials per condition. Is this correct?

I would greatly appreciate your help,
Sincerely,
Batiah Keissar, MA Student at University of Haifa