Infant Attention as a Joint Function of Visual Saliency and Parental Handled Actions

• In Event: Why Parent-Infant Interaction Matters: Impacts of Parent’s Multimodal Inputs on Infant’s Perceptual Experiences

Abstract

Visual exploration represents an early manifestation of motivation that provides a cornerstone for self-regulation and cognitive functioning. Literature on selective attention reveals a developmental shift from exogenous to endogenous control over visual attention during the first year of life (Colombo, 2001; Ruff & Rothbart, 1996). While attention is under development, parents provide references to support infants’ perceptual experiences. However, little is known about how visual selection emerges and develops in a free viewing context. The present study aims to address two questions regarding infants’ selection attention in the context of parent-child object play: (1) how does the saliency of infants’ scenes changes along with parental handled actions? and (2) how does the infant respond to the saliency change and handled actions and shift attention to the handled objects?

To address these questions, the present study observed a 5-minute-30-second free-flowing object play by a sample of 30 parent-infant dyads, including infants aged from 3 to 12 months (mean = 8.0 months, SD = 3.2). In the object play sessions, the parent was asked to play with the infant with a set of toys. We used a head-mounted eye tracker to record individuals’ moment-to-moment gaze behaviors. Trained coders annotated both parent’s and infant’s (1) gaze allocation to target objects and (2) each hand’s action onto objects (Figure 1A). The videos of head-mounted camera recordings were also fed into a saliency program (Itti et al., 1998; Itti & Koch, 2000) that employed Itti’s saliency algorithm (Figure 1B).

The preliminary analyses demonstrated the saliency change of infants’ visual scenes according to parents’ handled actions onto the objects. A linear mixed model with random slopes for dyads and trials suggested that the saliency of infant’s scene was significantly increased when the parent handled the object (ß = .010, p < .001) or the parent and infant handled the object together (ß = .021, p < .001). Moreover, a generalized linear mixed model with the random slope for dyads was used to examine the likelihood of infant object attention as a function of saliency change and object handling. The model revealed significant interaction effects of (1) saliency and parent object handling (ß = 2.25, p < .001) and (2) saliency and joint object handling between parent and child (ß = 1.83, p < .001). Figure 2 presents the predicted probability of infant object looking that infants were significantly more likely to attend to the object when it was handled by the parent or by the parent and infant together, and the likelihood of object attention was increased along with the enhanced saliency in the scenes.

The present study documents the contribution of visual salience in organizing infants’ attention and reflects the saliency change generated through parental handled actions. The present work offers an explanation of the role of parent object handling in nurturing an infant’s voluntary control and selective attention. Our ongoing analyses will further examine potential developmental changes in infant sensitivity to visual saliency in relation to parent’s object handling.

Authors

• Lichao Sun, University of Houston
  Presenting Author

• Hanako Yoshida, University of Houston