Dear SPM experts,
I'm analysing seed-based (4 ROIs) resting-state functional connectivity by using spm12 and especially DPARSFA toolbox (http://rfmri.org/DPARSF ).
The correlation coefficient values (Fisher-transformed (zFC map)) entered into second-level analyses indicate signed association values (positive values for positive association/correlation, and negative values for anticorrelations). In the 2nd level, I have a 2-by-2 within-subjects factorial design ANOVA, with factors A (two levels, A1 and A2) and B (two levels, B1 and B2). I am using a flexible factorial design in SPM12.
I have a significant difference between conditions for the following contrast:
Main effect of A (A2>A1, T-test) : -1 1 0 0 -0.5 -0.5 0.5 0.5
Main effect of B (B2>B1, T-test): 0 0 -1 1 -0.5 0.5 -0.5 0.5
Interaction A x B (F-test): 0 0 0 0 1 -1 -1 1
As I understand, the 2nd level positive/negative connectivity values do not reflect to the positive/negative correlation values, according to in my contrast. What is the correct way to distinguish between the positive and negative correlations and how can I display the results?
Another thing, do you prefer to remove the negative correlation maps from the individual zFC map before the 2nd level analysis (https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind1007&L=spm&F=&S=&P=538562 )?
A short clarification would be greatly appreciated.
All the best,
Arkan
|