Assessment of repeated displays: a test of possible mechanisms