Preferential looking is a foundational method in infant cognitive research, exploiting the fact that infants systematically look longer at some stimuli than others. Robert Fantz (1961) pioneered this approach, demonstrating that even newborns prefer patterned over plain stimuli and face-like over non-face configurations. The method has been extended into the habituation-dishabituation paradigm and the violation-of-expectation paradigm, which have been used to study virtually every aspect of infant cognition.
Paradigms
In the preferential looking paradigm, two stimuli are presented side by side and looking time to each is measured. Longer looking at one stimulus indicates discrimination and preference. In habituation-dishabituation, infants are familiarized with a stimulus until looking time decreases, then a new stimulus is presented. Recovery of looking time (dishabituation) indicates the infant discriminates the new from the old stimulus. In violation-of-expectation, infants watch events that either are consistent or inconsistent with a physical or psychological principle, with longer looking at inconsistent events interpreted as surprise.
The interpretation of looking time measures has been debated. Does longer looking at an impossible event really indicate conceptual understanding of the violated principle, or might it reflect simpler perceptual preferences or novelty detection? This debate has motivated the development of complementary methods including reaching measures, anticipatory looking, and neural measures (EEG/ERP), providing converging evidence about infant cognitive abilities.