split the analysis per block?

Hi you all,

I was wondering if there was any way to split my analysis per block so that I can see the changes in activation over time. Let me explain more: I just read a paper showing " dramatic decrease in ACC activity and an increase in DLPFC activation when the first half of the trials were compared with the second half of trials ([Erickson et al., 2004]" and explained by practice effect.

In my experiment I have four blocks of a numeric Stroop task in an event related design (randonly presenting congruent, and incongruent trials). I already analyzed it to get the Incongruent > Congruent contrast across blocks. But now I was wondering if there is anyway I could split the data in half to see how activations change with practice.
I guess the easy way would just be to run afni proc py for block 1&2 vs. block 3&4 but if there is a smarter way to do that that would be great .Thank you!