Abstract
Cues such as directed gaze, facial expressions, and manual gestures provide rich information that facilitates social interactions. Using looking time methods, previous studies have found that infants prefer faces to non-face stimuli (Mondloch et al., 1999; Morton & Johnson, 1991), discriminate biological from non-biological motion (Bertenthal et al., 1987; Simion et al., 2008), and are sensitive to others’ gaze and eye contact (Hood et al., 1998; Farroni et al., 2002). These, however, have been examined in isolation. Here we present a new eye tracking paradigm for dynamically studying infants’ and adults’ visual attention to competing social cues while viewing naturalistic action sequences, to more precisely examine how and when visual attention shifts and how this varies with age. Sixteen video stimuli (average duration = 24 sec) were produced with five female actors performing infant-directed actions (e.g., pouring into a mug, placing a bow on a box; see Figure 1). Eight- and twelve-month-old infants and adults (N = 62) were shown these videos while gaze was measured with a Tobii corneal reflection eye-tracking system. Areas of interest dynamically defined the actor’s face, hands, and objects (see Figure 2). The actor’s social cue events (e.g., smiling, speaking, eye contact, and reaching for, grasping, and holding the objects) were observationally coded. The data reveal strong correlations between the two groups of infants and adults, though infants’ fixations lagged behind adults’ by about 250ms (see Figure 3). Furthermore, infants responded to triadic relations differently than adults, and whereas12-month-old infants attend more to actions and objects, 8-month-old infants consistently attend more to the actors’ faces. Additional analyses reveal adults are more sensitive to observationally coded social cues (e.g., smiling, speaking, and eye contact). These results highlight the importance of fine-grained spatiotemporal analyses and suggest development of sensitivity to social cues beyond the first year.
Meeting abstract presented at VSS 2012