Predicting children's media use in the USA: differences in cross-sectional and longitudinal analysis.

The purpose of this paper is to examine the predictors of children's media use in the USA, comparing cross-sectional and longitudinal analyses. Data come from Waves I and 2 of the Child Development Supplement (CDS-I; CDS-II), a nationally representative sample of American children aged 0-12 in 1997 and 5-18 in 2002. Twenty-four hour time use diaries are used to assess children's time spent with media (television, video games, computers, and reading). Predictors examined include socio-demographics, neighbourhood quality, family factors, and other media use. Ordinary least square (OLS) multiple regressions were performed by three age groups (preschoolers, early school age, and preadolescence). The findings suggest that neighbourhood quality, parental limits and family conflict are significant predictors of children's media use within time or over time, but the significance depends on the type of media and child's developmental stage. In addition, children's television viewing and reading habits are formed early in life and reinforced over time. This study is among the first to provide empirical evidence for the effect of early contextual factors on the life course of children's media use from a developmental perspective.