Hi Anderson,
First off, thanks for all of the support you provide - it's incredibly helpful!

I had a question about your comment from here:

https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind1708&L=FSL&F=&S=&P=231860

Where the question was:

> Question 4: What if I have 2 groups and I want to examine whether the relationship >between M1 and M2 is significantly different across the 2 groups? Would the design >matrix be 2 columns for group membership (1s and 0s)? The contrast is 0 0 -1 1?  

And your response was:
>This would require 2 voxelwise EVs. The input would be M1, the first EV would be M2, >the second EV would be M2 times the +1/-1 that defines group membership, and a third >EV would be the intercept. The second EV would investigate the interaction, which is >what you want. The contrasts would then be [0 1 0] and [0 -1 0].


It's unclear to me what exactly should be done for creating M2. For comparing two groups, I'm used to coding my group membership in the design as:

1 0
1 0
0 1
0 1
etc.

Could you clarify what is meant by the +1/-1 in terms of "defining group membership"? I naively thought that it might be something like:

1 -1
1 -1
-1 1
-1 1
etc.

but this seems to be incorrect, especially since M3 is then being used to code the intercept, so clearly I must be wrong here. 

If you could provide a toy design matrix like those above, I would be very grateful! I am assuming also that M2 would then be multiplied by one of the EVs. I guess the easiest in this case is to do something like fslsplit, multiply each volume by its corresponding value, and then remerge? Or have I completely misunderstood what you meant?

Thank you again very much!

Best,
Joe