Dear Russ,
> Can someone from the SPM group clarify how the resulting numbers are
> scaled when one performs an event/epoch-related response plot? It
> would be nice to be able to interpret them in terms of some actual
> quantity (like % signal change) but I am guessing that they are not
> scaled that way.
By default the data are normalized to a grand whole-brain mean of 100.
This means that modeled effects are expressed in adimensional units
corresponding to percent whole brain signal.
I hope this helps - Karl
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|