Third, Connellan et al. (2000) did not discuss critical controls against experimenter bias. Because newborn infants cannot hold their heads erect, their visual preferences are influenced by the way in which they are positioned and supported; because one of the two stimuli was a live, expressive person, preferences also could be influenced by that person’s behavior. Baron-Cohen (2005a) has indicated that the experimenters attempted to minimize bias, but a replication with more stringent controls would be desirable.