Accidental injuries in childhood: Comparison of the contributions from parents and childre

Accidents are responsible for more of the fatal injuries of children between the ages of 1 and 14 than the next six leading causes of death combined and the frequency of nonfatal accidents is equally substantial. Previous research has focused predominately on two groups of variables that affect children's accident risk: characteristics of children themselves and the characteristics of their parents and families. Two models, derived from the findings of previous research, were compared for their relative ability to account for differences in children's histories of medically-attended injuries. One model was composed of the child characteristics of activity level, behavior, and motor coordination, whereas the other model was composed of the parent and family factors of stress, stability, and locus of control. Two hundred five mothers of children between the ages of 3 and 12 years old completed a survey assessing these variables. The children of the mothers varied with respect to the number of medically-attended nonintentional injuries they had sustained during the past year. The major hypothesis of this study was that the model composed of parent and family variables would more accurately account for differences in the accident histories of children than the model composed only of child characteristics. Thus, it was predicted that children's histories of medically attended nonintentional injuries are most strongly influenced by parent and family variables. The data generated from this study provided support for this hypothesis. The second hypothesis was that there would be a developmental trend in the ability of the two models to account for differences in the accident histories of children. Specifically, differences in the accident histories of the younger children would be best explained by the Parent model; while for older children, their characteristics would be more important in accounting for differences in their accident liabilities. This hypothesis was not supported by the data. Lastly, the hypothesis that the combined model composed of all variables would not significantly improve prediction of all children's accident histories was tested and supported by the data. The implications of these results with regard to the design and implementation of prevention programs is discussed.