Mixing incompatibly mapped location-relevant trials with location-irrelevant trials: effects of stimulus mode on the reverse Simon effect