one object that tested their ability to learn part locations, and one object that assessed their fatigue and boredom. The objects that tested infants’ learning of the second-order correlation both had a novel body: one object was connected to the two parts that were previously associated with the same body (the consistent test object), and another object was connected to the two parts that were previously attached to different object bodies (the inconsistent test object). Longer examining—characterized by clearly focused looking—of either the consistent or inconsistent test object could only result from infants learning the second-order correlations in the familiarization stimuli. Experiment 2 used a similar design to address whether 9-month-olds’ second-order correlation learning is facilitated by paired presentation of the test stimuli. Experiment 3 was designed to examine whether 7-month-old’s second-order correlation learning is affected by making the familiarization stimuli more distinct, and Experiment 4 was devised to explore the mental representations that 11-month-olds form during the process of second-order correlation learning.