Hi again,
the deconvolutions came out well! I have a question about a couple of regions that show a clear DEACTIVATION (the hemodynamic response is beefy, but negative-going). The paradigm is the usual: condition1...4, plus some fixations during the ISIs (sequence was optimized etc etc). The noise model is a second order pol. I'm wondering whether this "deactivation" can be interpreted in any meaningful way. One of the issues I have is that all conditions are negative at this location. Say, condition 1 is larger and more negative than condition 2. I can explain this by saying that condition 2 is deactivating this region more than condition 1. But how cannot discount the possibility that the entire deactivation is due to some hydraulic phenomenon (blood being stolen from this region with an overlapping condition effect going in the opposite direction). Furthermore, deactivation relative to what? Since the fixation epochs are not modeled explicitly, I'm not sure... Is it relative to the average of the noise model for the entire scan? Thanks for any suggestions/thoughts.