Mixing location-irrelevant and location-relevant trials: Influence of stimulus mode on spatial compatibility effects