| The study examined the differential influence of school transitions on the reading skills of poor and non-poor children from kindergarten through fifth-grade using nationally representative data (ECLS-K). The study employed a latent growth curve model to estimate the impact of school moves on the initial status and slope of the IRT-scaled reading scores (dependent variable) of children by poverty status across time, adjusting for random time-invariant and nonrandom time-variant covariates. The random time-invariant covariates were age, gender, and race. The nonrandom time-variant covariates were school mobility, school level SES, mothers' level of education, and types of family structure at each time-point. Six models and a sensitivity analysis were used to assess the impact of school mobility on the reading scores of poor and non-poor children. The results revealed a negative impact of school mobility (about 3 points) on the IRT-scaled reading scores in kindergarten for poor children. Other model covariates (such as race and mothers' level of education) were found to be more influential on reading skills than school moves. The results showed small to medium effect size differences between non-poor Black children and non-poor children of other races; except for Black children, the results showed generally small to medium effect size differences between poor and non-poor children within the same race from first-grade through fifth-grade. |
