Early-emerging gradients in children’s eye movement times in low-resource settings

Abstract

A growing body of evidence shows that early child development varies along a socioeconomic gradient, with poorest outcomes observed in chil- dren from the most disadvantaged environments in low-income countries. Most of this evidence comes from studies using caregiver reports or expert judgments of child behavior. We used eye tracking to examine whether and at what age direct measures of elementary visual behav- ior begin to differ in children experiencing different levels of poverty in rural South Africa. Saccadic reaction times (SRTs) were assessed at child age 7, 17, and 36 months. Mean SRT was inversely associ- ated with household wealth, with slower SRTs observed in children from less wealthy households. This gradient was detectable from 17 months onward, tended to increase in magnitude by age, was not explained by differences in screen time, and was comparable to gradients seen in other developmental outcomes. The results point to pervasive socioeconomic gradient effects in early child development, including objective measures of behaviors that are foundational to early interactions and learning.

Evaluation of a community health worker home visit intervention to improve child development in South Africa: A cluster-randomized controlled trial

“…Lower SRT is indicative of advanced underlying neural maturation and white matter integrity [53,54]. In an analysis we present elsewhere, higher household wealth was strongly associated with lower SRT in the study sample [55]. Previous studies have shown that lower SRT measured during infancy is predictive of stronger cognitive and executive function skills expressed later in childhood [56,57].…”

Is Reliability of Cognitive Measures in Children Dependent on Participant Age? A Case Study with Two Large-Scale Datasets

“…In addition, as an aggregate "study-level" phenomenon, we measure correlations and their CI95s between SES scores of infant households and infant SRTs for the different age bins. This is since earlier research has reported a reduction in SRTs with increasing household wealth on the same data (Leppänen et al, 2023).…”

“…As expected (e.g., Alahyane et al, 2016), the SRTs systematically decrease with age (r = -0.57, p < 0.001). In addition, household wealth (SES) and SRTs are inversely correlated in the 17-and 36-month age groups, as reported by Leppänen et al (2023).…”

“…As the second dataset, we utilize SRT data collected from Greater Tzaneen area within Mopani District, Limpopo Province, South Africa, as originally described in Leppänen et al (2023). The data consists of responses from an SRT experiment administered to infants at 7-, 17-, and 36-month checkpoints in a longitudinal manner.…”

“…Household wealth was estimated based on a checklist of 29 assets that a household might own, subjected to principal component analysis as described by Filmer and Pritchett (2001). The reader should see Leppänen et al (2023) for complete details of the experimental setup, participants, and collected data.…”