Yes, just add to the design an EV with those events. I assume the random rest events were random for the subject, but ideally you'd know when they happened.
For a rapid event-related design in which there are two types of stimuli and also random rest events, is it possible to model response solely for one stimulus type? If yes, how?
