Hi Alfonso,

I am wanting to extract raw connectivity values from ROI-ROI analysis. From my understanding, the Z table that is provided for each subject is a fisher transform of the pearson correlation. but some of the values in my table are above 1. I am wondering why that is and if it would be wrong to plot such values.

thanks

Hi Nancy,

Yes, you are right, if you are running RRC first-level analyses with default options (using correlation measures) then those Z matrices contain the Fisher-transformed correlation coefficients, and it is perfectly fine than some will exceed 1 as the mapping/association between r and Z values is:

r = tanh(Z)

so, for example, a Fisher-transformed value Z=1 corresponds to a Pearson correlation coefficient value r = 0.7616 (and of course you can either plot those values directly or first transform them to correlation coefficients and then display those, both are perfectly valid/meaningful approaches)

Hope this helps
Alfonso
Originally posted by nancymugisha: