050 - Enduring effects of screen time on language development

Publication Number: 50.5094

Fabeha Husain, national university of singapore, Singapore, N/A, Singapore; Evelyn Law, National University of Singapore, Singapore, N/A, Singapore

Background: Longitudinal associations between screen use and poor language development have been reported using regression and cross-lagged panel models. This study examined the effects of early screen time on language skills using the enduring effects model. This model tests whether effects of early exposures persist over time, the effects fade (i.e., revisionist model), or outcomes are better predicted by contemporaneous experiences.

Objective: To examine whether early screen use has enduring effects on language development at ages 2, 4, and 4.5 years.

Design/Methods: Data were from the Growing Up in Singapore Towards healthy Outcomes (GUSTO) prospective cohort. At age 1, 2, 4, and 4.5 years, parents reported the amount of screen use of the child, weighting weekdays and weekends [(time on weekdays × 5 + time on weekends × 2) / 7] to obtain an average daily screen time (hours/day). Language skills were assessed using the Bayley Scales of Infant and Toddler Development, Third Edition (BSID-III), Peabody Picture Vocabulary Test, Fourth Edition (PPVT-IV), and Verbal Knowledge from the Kaufman Brief Intelligence Test, Second Edition (KBIT-II) at ages 2, 4 and 4.5 years, respectively. Standard scores (Mean=100, Standard Deviation [SD]=15) were used. Covariates were child sex, monthly household income and maternal education. Structural equation models (SEM) with a robust maximum likelihood estimator were used to compare model fits of the enduring effects, equal enduring effects and revisionist models under three conditions: without covariates, with covariates, and with additional exposure timepoints (Figure 1a-c).

Results: The mean (SD) daily screen time at age 1 year was 2.0 (1.9) hours per day and was negatively correlated with all language scores (r= -0.274 to -0.331). Higher household income and maternal education were both associated with lower screen time and higher language scores at all ages (Cohen’s d: 0.387 to 0.836). Hence, household income and maternal education were included as covariates in the SEM models. The enduring effects model was the best fitting among the three conditions (Table 1). All associations between screen time and language outcomes (i.e., a, b1, b2, c1, c2 paths) were statistically significant (p < 0.05) in the enduring effects models.

Conclusion(s): Our findings suggest that there may be enduring effects of early screen time exposure on language development throughout childhood. This may add to extant literature on the longitudinal associations between early screen use and language outcomes in children.

Figure 1 a-c.


Table 1.