Online data collection with infants raises special opportunities and challenges for developmental research. One of the most prevalent methods in infancy research is eye-tracking, which has been widely applied in laboratory settings to assess cognitive development. Technological advances now allow conducting eye-tracking online with various populations, including infants. However, the accuracy and reliability of online infant eye-tracking remain to be comprehensively evaluated. No research to date has directly compared webcam-based and in-lab eye-tracking data from infants, similarly to data from adults. The present study provides a direct comparison of in-lab and webcam-based eye-tracking data from infants who completed an identical looking time paradigm in two different settings (in the laboratory or online at home). We assessed 4-6-month-old infants (n = 38) in an eye-tracking task that measured the detection of audio-visual asynchrony. Webcam-based and in-lab eye-tracking data were compared on eye-tracking and video data quality, infants’ viewing behavior, and experimental effects. Results revealed no differences between the in-lab and online setting in the frequency of technical issues and participant attrition rates. Video data quality was comparable between settings in terms of completeness and brightness, despite lower frame rate and resolution online. Eye-tracking data quality was higher in the laboratory than online, except in case of relative sample loss. Gaze data quantity recorded by eye-tracking was significantly lower than by video in both settings. In valid trials, eye-tracking and video data captured infants’ viewing behavior uniformly, irrespective of setting. Despite the common challenges of infant eye-tracking across experimental settings, our results point toward the necessity to further improve the precision of online eye-tracking with infants. Taken together, online eye-tracking is a promising tool to assess infants’ gaze behavior but requires careful data quality control. The demographic composition of both samples differed from the generic population on caregiver education: our samples comprised caregivers with higher-than-average education levels, challenging the notion that online studies will per se reach more diverse populations.

While interpersonal synchrony is regarded as a defining characteristic of early social exchanges between infants and their caregivers, the mechanisms facilitating synchronous interactions are poorly understood. The goal of the present study was to examine the relationship between maternal playful singing and interpersonal synchrony of the mother-infant dyad. Overall, 56 mothers and their 4-month-old infants were observed during natural interactions that were then micro-coded for maternal and infant social behaviours as well as maternal singing and rhyming in the context of social game routines. Results showed that 38 dyads spontaneously engaged in social game routines. In these dyads, both playful singing and rhyming were positively associated with dyadic gaze synchrony, while only playful singing was also positively correlated with affect synchrony of the dyad. These findings suggest that rhythms, in general, may have important implications for the establishment of interpersonal synchrony in infant-caregiver dyads. However, musical rhythms seem to be particularly emotionally-salient and thus attune both interactional partners to the affective content of their social exchanges.

Much recent work has shown that children actively shape their learning progress by choosing what, when and from whom they want to learn. However less is known about whether children are sensitive to gaps in their knowledge, and whether they selectively solicit information about items they previously indicated having a knowledge gap about. In a cross-situational word-learning task, we asked 5-year-olds, 6-9-year-olds and adults to estimate their knowledge of newly learned word-object associations. We then examined whether participants preferentially chose to hear the labels of objects they reported not knowing the label of and whether such uncertainty-driven sampling behavior led to improved learning. We found that participants in all age groups were sensitive to gaps in their knowledge of the word-object associations, i.e., were more likely to say they had correctly indicated the label of an object when they were correct, relative to when they were incorrect. Furthermore, 6-9-year-olds and adults were more likely to sample objects whose labels they reported not knowing, while 5-year-olds were equally likely to sample objects whose labels they reported not knowing and knowing. In other words, older children and adults displayed sampling behavior directed at reducing knowledge gaps and uncertainty, while younger children did not. However, while all age groups showed evidence of having learned the word-object associations, participants who displayed more uncertainty-driven sampling behavior were not more accurate at test. Our findings underscore the role of uncertainty in driving 6-9-year-olds’ and adults’ sampling behavior and speak to the mechanisms underlying previously reported performance boosts in active learning.