Evaluating Screen Time and Child Development

To the Editor On January 28, 2019, the Madigan et al study was published in which they “support the directional association between screen time and child development.” The article received considerable public attention, but its conclusion is problematic. First, all between-person associations were not significant, and the within-person cross-lagged associations were negligible (β , −0.08 and −0.06). Conventional estimates of effect size (eg, Cohen d, R2) were not provided, but assuming that the variables were standardized, the β coefficients equal Pearson r. De facto, the effect sizes, which are the extent to which the predictors (ie, screen time) contribute to the expl anation of the variance in the dependent variables (ie, developmental outcomes), approach zero (0.0036 ≤R2 ≤0.0064).
Source: JAMA Pediatrics - Category: Pediatrics Source Type: research