Developmental changes in infants’ and children’s attention to faces and salient regions vary across and within video stimuli.

Abstract

Visual attention in complex, dynamic scenes is attracted to locations that contain sociallyrelevant features, such as faces, and to areas that are visually salient. Previous work suggests that there is a global shift over development such that observers increasingly attend to faces with age. However, no prior work has tested whether this shift is truly global, that is, consistent across and within stimuli despite variations in content. To test the global shift hypothesis, we recorded eye movements of 89 children (6 months to 10 years) and adults while they viewed seven video clips. We measured the extent to which each participant attended to faces and to salient areas for each video. There was no evidence of global age-related changes in attention: Neither feature showed consistent increases or decreases with age. Moreover, windowed analyses within each stimulus video revealed significant moment-to-moment variations in the relation between age and each visual feature (via a bootstrapping analysis). For some time windows, adults looked more often at both feature types compared to infants and children. However, for other time windows the pattern was reversed-younger participants looked more at faces and salient locations. Lack of consistent directional effects provides strong evidence against the global shift hypothesis. We suggest an alternative explanation: Over development, observers increasingly prioritize when and where to look by learning to track which features are relevant within a scene. Implications for the development of visual attention and children's understanding of screen-based media are discussed.

Developmental changes in natural scene viewing in infancy.

“…We chose to use these saliency models for several reasons. First, they are commonly used in understanding infants’ fixation (Kadooka & Franchak, 2020; Kelly et al, 2019; Mahdi et al, 2018; Simpson et al, 2019). Thus, a systematic investigation of how well these models predict infants’ fixations will be useful for the interpretation of these other findings.…”

Age differences in orienting to faces in dynamic scenes depend on face centering, not visual saliency

“…First, rates of looking at salient areas and faces are highly dependent on the particular video stimuli. When comparing across a wide range of ages (6 months to 12 years) and across a large, diverse set of video stimuli, there was no evidence for consistent age differences in either looking at salient areas or at faces across video clips (Kadooka & Franchak, 2020). Although a few clips showed age trends, age differences were not evident for most stimuli.…”

“…And although centering is a strong cue for adults’ orienting to faces in videos (Xu et al., 2018), no study has tested whether centering relates to face looking in infants and children. Previous research found considerable variability in rates of looking at faces, rates of looking at salient areas, and the saliency of faces across video stimuli (Frank et al., 2012; Kadooka & Franchak, 2020; Stoesz & Jakobson, 2014; Wass & Smith, 2015). Accordingly, it was important for the current investigation to test saliency and centering cues for face looking across a large stimulus set with multiple faces that vary in appearance and location over time.…”

“…To better understand these interactions, we can investigate how saliency might serve as a cue to help guide attention toward faces. Rather than testing whether observers look at faces versus salient areas (e.g., Franchak et al., 2016; Frank et al., 2009; Kadooka & Franchak, 2020; Rider et al., 2018), we can instead ask whether observers look more often at a face when it is more salient compared to when it is less salient . Salient faces might help scaffold infants’ scene viewing through the convergence of bottom‐up (e.g., visual appearance) and top‐down (e.g., role in the narrative) cues.…”

“…If we find age‐related changes in how face looking depends on saliency and/or centering, it is important to determine whether those changes in attention are specific to face looking or reflect general attention development. Although the past study using the data set (Kadooka & Franchak, 2020) found that there were no general age‐related changes in how frequently infants, children, and adults looked at salient areas or at faces, the role of centering was not tested. Thus, we conducted a second “overall looking analysis” to assess whether there were age‐related changes in looking at centered regions, salient regions, and faces.…”