| The purpose of this study was to examine the stability of relational aggression among females in grades 3, 4, 5, and 6, and longitudinal relationships between relational aggression and the variables of physical aggression, perceived popularity, sociometric status, pubertal onset, social skills, peer support, friendship quality, and parental support. Hierarchical linear modeling was used to analyze data in this study. In order to explore the influence of predictors on relational aggression over time, level-1 analysis modeled the within-person trajectory of relational aggression with repeated measures of relational aggression and the time-varying covariates (physical aggression, social skills, friendship quality, and parental support) for each student. Level-2 analysis included time-invariant variables (pubertal onset, sociometric popularity, perceived popularity, and peer group).;Results indicate that there are differences between and within individuals in relational aggression. The interclass correlational coefficient for the unconditional means model indicates that approximately 32.6% of the variance in relational aggression lies between individuals. The results of the unconditional linear growth model indicate yearly growth in relational aggression was -.0195 points per year. The fixed effect representing initial relational aggression in the unconditional linear growth model was found to be significant at the p < .001 level. Results of the full model indicated statistically significant results at the p < .001 level in initial relational aggression for sociometric status. The relationships between time-varying covariates of physical aggression, social skills, and friendship quality and yearly growth in relational aggression were also statistically significant. The calculation of R˜2 statistics show that the addition of variables accounted for approximately 29.3% of the variation in initial relational aggression and for 56.9% of the variation in growth in relational aggression. A positive correlation between initial relational aggression and yearly growth rate in relational aggression (r = .863) indicates that after controlling for all covariates, participants who have higher levels of relational aggression at third grade have higher annual growth in relational aggression. Comparisons of deviance statistics suggest that each subsequent model fits the data better than each previous model, with the full model having the best fit to the data. 2... |
