Contingent behavior during caregiver-child interaction improves the quality of word learning opportunities.

We examined how the contingency of caregiver-child behavior influences child attention to named objects ("targets") during free-flowing toy play. Using gaze data collected from head-mounted eye tracking, we operationally categorized caregiver naming instances into three types based on the contingency between child attention and caregiver speech: (1) child-led contingency instances, where children were attending to the target prior to the onset of caregiver naming, (2) caregiver-led contingency instances, where children were not attending to the target at the onset of caregiver naming but switched their attention to the target during the utterance, and (3) non-contingent instances, where children did not attend to the target object at all during caregiver naming. We found that target attention for contingent instances was higher than for non-contingent instances. Further, target attention was higher for child-led instances than for caregiver-led instances. In child-led instances, caregivers tended to name the object that was in children's hands, and in caregiver-led instances, children tended to orient their attention to the object in their caregivers' hands. These results show that contingency of caregiver-child behavior has crucial implications for the quality of word learning input, and that object holding serves as a powerful cue for creating contingency in caregiver-child social interactions.