Interpreting infant looking: the event set x event set design

Theory, data, and mathematical models presented suggest that perceptual processing may be crucial in young infant cognition. Prior results indicating early or innate physical knowledge are reinterpreted. Assumptions that young infants use higher level cognitive processes to infer, reason, believe, and so on are challenged in favor of perceptual processes and the effects of novelty and familiarity. The 2-test habituation design that compares looking at the "possible" with looking at the "impossible" and the problems of that design are considered. The authors' approach, based on regression analysis of Event Set X Event Set factorial designs, eliminates those problems, refines gauging the contribution of various variables, quantifies these contributions with standard parameter estimation, unconfounds the crucial variables, and tests which variables are responsible for looking time differences. Data are presented that support the perceptual processing perspective. Application of the new design to 2 seminal studies of infant cognition are suggested.