Using a confidential Canadian dataset of children and youth (National Longitudinal Survey of Children and Youth), I have provided empirical evidence of the school performance, bullying behavior, and language immersion of children in four chapters of the Ph.D. thesis.
In the first chapter, the academic performance of children of immigrants is compared with that of their classmates of Canadian-born parents. The comparison starts when children are in kindergarten and continues until they grow up to become adolescents.
In the second chapter, the bullying behavior of children is explored.
This chapter focuses on the identification of causality between parental control and children’s bullying behavior, which is generally under-investigated in the existing literature. First, we build a theoretical model to capture the strategic dependence of children’s bullying behavior and the corresponding parental control. Then, we employ conditional fixed effects logistic estimation to test the theoretical conclusions. The empirical results support our hypothesis that stricter disciplinary measures taken by parents are more effective in deterring the child from bullying when all the other
factors are held constant. The causality is carefully justified by making great efforts to account for all possible identification issues.
Chapter 3 studies the children’s bullying behavior in a dynamic scenario by answering the question of when is the best time to stop bullying. Results from the semi-parametric propensity score matching suggest that early bullying detection and intervention contributes to a positive suppression effect on it.
In the last chapter, we provide empirical evidence on who are in French immersion programs and who are more likely to drop out of French immersion.
Results from the two-stage least-squares estimation indicate that children with higher reading ability are more likely to enter French immersion programs. Both simple logistic estimation and duration analysis