The nature of the relationship between poverty and children's aggression during elementary school

The present study investigates whether the degree and timing of poverty affected developmental change in teacher-rated aggression in a sample of approximately 400 low-income children followed from kindergarten through fifth grade. Children living in poverty are more likely than their peers who are not poor to have a host of short- and long-term difficulties, including behavioral problems (Duncan & Brooks-Gunn, 1997). The effects of poverty on children's outcomes may depend, in part, on when children experience poverty as well as the degree of poverty they experience. Although a significant amount of research has documented differences between low, middle-, and high-income children regarding their development of social behaviors, there has not been as much research examining differences within the low-income group. In addition, growth curve analysis using Hierarchical Linear Modeling has rarely been used to examine both individual trajectories of aggressive behavior during elementary school as well as the influence of the degree and timing of being poor on these trajectories. Results indicated that being raised in a family more than half below the federal poverty threshold at kindergarten predicted a significantly higher rate of change in mean aggression through fifth grade. Also significant predictors of the rate of growth of aggression were being male and maternal marital status. Further analyses indicated that, whereas children's mean conflictual relationships with teachers did predict differences in children's aggression ratings in kindergarten, neither mean levels of teacher-child conflict nor the interaction between poverty and teacher-child conflict accounted for differences in the rates of change in aggression across the grades when taking into account other child and family background variables. In addition, the interaction between children's close relationships with their teachers and children's poverty level was a marginally significant predictor of differences in children's mean aggression ratings in kindergarten but not of differences in the rates of change in aggression across the grades.