Visual Attention to Child-Directed and Observational Contexts in US and Yucatec Mayan Households

• In Event: 1-099 - Factors that promote third-party learning in young children

Integrative Statement

Laboratory tasks have demonstrated that child-directed interactions exert proximal effects on infants’ visual attention that may support social learning (e.g., Cooper & Aslin, 1990). However, little is known regarding how child-directed vs. observational contexts capture attention in natural environments. Moreover, prior work has suggested that the relative importance that directed interactions have may vary across cultures: directed interactions may be less critical for shaping attention in rural Mayan communities because children are expected to learn from their ongoing observations of others (Rogoff et al., 1993). The current study investigated 18-month-old infants’ attention to natural directed vs. observational situations in in two cultural contexts: a US city and a Yucatec Mayan village. In addition, the study investigated how behavioral cues indicating directedness may relate to infant attention.

Hour long video was collected in visits to the homes of 18-month-old infants and their families (12 Maya, 12 US). Families were asked to go about activities as they normally would. Videos were coded for skilled actions performed by individuals around the child, defined as an actor using an object in a conventional way to accomplish a goal. Each segment of video containing an action was coded for if the actor used behavioral cues indicating directedness to the target child (joint eye-gaze, motionese, directed gesture). Segments that contained any directed cue were categorized as Directed Actions. Segments that did not contain any of these behavioral cues were categorized as Observed Actions.

Base rates of actions were different across cultures (MedianUS=85, range = 45-126; MedianMaya= 29, range = 12-124; t(21) = 3.19, p <0.01), However, the proportion of Directed and Observed actions did not differ (US proportion Observed = 0.85, Direct = 0.15; Maya proportion Observed = 0.82, Direct = 0.18). To assess whether children in each culture attended more to Directed or Observed Actions, we ran a linear regression with Duration of Attention to skilled actions as the outcome variable and Duration of skilled actions and whether the action was Directed or Observed as explanatory variables. Results showed that across cultures children attended longer to Directed vs. Observed actions (US: b = 9.40, p <0.001; Maya: b = 3.82, p < 0.05, see figure 1). Next, we asked whether children attended longer to Directed Actions that had more Directed Cues. We performed a linear regression using Duration of Attention to Directed Actions as the outcome variable and Duration of Directed Actions and the amount of Directed cues in each Directed Action as explanatory variables. The results show that for both cultures, children attend longer to Directed Actions that have more behavioral cues (US: b = 8.36, p <0.05; Maya: b = 10.23, p < 0.05, see figure 2).

Thus, in everyday environments, directed contexts are better able to capture infant attention as compared with observational contexts, and infant attention relates to the number of behavioral cues actors use indicating directedness. This is true even for infants growing up in a cultural community where learning from observation has been described as expected and valued.

Authors

• Laura Shneidman, Universidad Nacional Autónoma de México
  Presenting Author

• Nicole Burke, University of Chicago
  Non-Presenting Author

• Amanda Woodward, University of Chicago
  Non-Presenting Author