Home environmental factors associated with screen time among Japanese preschool children: a cross-sectional secondary analysis using exploratory structural equation modeling

Screen time in early childhood is increasingly recognized as a behavior shaped by family and social environments. This study examined the structural associations of television viewing and electronic game playing with multiple home environmental factors among Japanese preschool children. This cross-sectional secondary analysis used data from the fifth wave of the Longitudinal Survey of Newborns in the 21st Century, conducted in 2014 when the children were 4 years and 6 months old. The complete-case analytical sample comprised 19,577 children from two-parent households. Screen time was modeled as a latent variable indicated by television viewing time and electronic game-playing time. Exploratory factor analyses generated indicators of parenting-related perceptions and paternal and maternal involvement. We constructed and compared theory-informed exploratory structural equation models (SEMs) and selected the final model based on fit, estimation stability, theoretical interpretability, and parsimony. More than 30% of children watched television for at least 3 h per day, whereas 63.9% did not usually play electronic games. Among three alternative SEMs, the model using Factor 2 indicators of parental involvement had the most favorable overall fit (CFI = 0.907, AGFI = 0.962, RMSEA = 0.047, SRMR = 0.0407). The final model explained 50% of the variance in screen time. Socioeconomic status (SES) had the largest standardized total association with screen time (− 0.550). Parenting-related subjective appraisal and paternal and maternal involvement also had model-implied associations with screen time, but their standardized magnitudes were smaller. Preschool children’s television viewing and electronic game-playing time were structurally associated with household socioeconomic conditions, parenting-related perceptions, and paternal and maternal involvement. This exploratory cross-sectional SEM cannot establish causality, but the findings identify associations among household resources, caregiving burden, parent–child involvement, and screen-based behavior that warrant longitudinal investigation. Because the data reflect television- and game-based screen use in 2014, the findings provide a historical reference rather than an estimate of contemporary total screen use.