Hi Alfonso, 

I would like to follow up on my question. I've tried some things but still couldn't figure out what to do. 

Again, my intention is to plot and run post-hoc analysis in another software. If only "AllSubjects" is selected in "Subject Effects", and "Any effects (F-test)" in "Between-conditions contrast" (see figure below), I can get the connection values for each condition and subject by clicking on "plot effects". The subject- and condition-specific connections are exported to MATLAB workspace. 

Upon clicking on "plot effects", I also visually get the averages and standard deviations from each condition-specific connection:

However, if also the covariates are selected in "Subject Effects", and the "Between-subjects contrast" is set to [1 0 0], I get the same values as before in the MATLAB workspace when I click on "plot effects" (here, I understood that "age" and "hand" are removed from "AllSubjects" and the ANOVA is made with a new "AllSubjects" corrected by these confounding variables): 

Although the subject- and condition-specific connections are the same as before in MATLAB, the "connectivity effects"  window show something else this time: 

By checking the code in "conn_glm.m", I found out that what is being shown now is the beta for each condition (line 96), not the subject-specific connection like before. 

I thought that the variable E (line 94: E=Y(:,:)-X*B) was the connection values corrected by the confounding variables I was looking for, but their average is zero for each condition.  

Is there a way to get the corrected subject- and condition-specific connections, corrected by the confounding variables, with CONN? Or did I misunderstood something? I hope that I made the question clear.

I know that alternatively I could run all the post-hoc analysis and plotting based on CONN, or drop the covariates... But I think it would be great if I could get the correct values from CONN (Another alternative  of course would be doing the analysis (GLM and ANOVA) in R instead of CONN, but I would need to see if I can reproduce it)

Cheers,

Gustavo