Hi Pat,

The baseline condition is used somewhat interchangably
with the rest condition, at least in task analysis. It
represents the periods of time that do not correspond to
any stilmulus events. In AFNI, that is an implicit part
of the regression model (i.e. one does not specify such
times), and which is modeled separately per run (this is
a key point for you, it seems).

Saying that you have "4 conditions and rest in between"
makes me a bit nervous. It might be good to be very
precise about your event timing. Also, 4 conditions
plus rest in between sounds like 4 predictors, not 5.

- rick