In this study, researchers from the University of Manitoba, McGill University, and the University of Western Ontario looked at extent to which injuries to children in cases of reported child physical abuse could be predicted by characteristics of the children, the characteristics of the perpetrators, or their socio-economic characteristics. The study tested the assumption that physically injurious incidents of child physical abuse are qualitatively different from those that do not result in injury. Working with a Canadian Incidence Study data set of 8,164 substantiated cases of physical maltreatment by at least one birth parent in which an investigative case worker deemed there had been inappropriate punishment, the researchers constructed various models to try to predict injury. Injury was initially recorded as physical harm in various categories but, because there were relatively few injuries, this study only looked at predicting whether a child had been injured (i.e., type of injury could not be predicted).
Results showed more than half of the abuse incidents did not result in physical injury to the child, and that none of the potential predictors (child age, gender of perpetrator, child functioning, parent functioning, economic stress and social stress) predicted injury to the child. The findings suggest that injurious physical abuse cannot be distinguished from non-injurious physical abuse on the basis of the personal characteristics or circumstances of the child or the perpetrator. Practice implications: Child welfare workers are often called upon to predict whether a child is at risk of injury when making decisions as to whether to intervene in situations of potential maltreatment. The findings of this study suggest that in substantiated cases of parental abuse in the context of punishment, injury cannot be predicted by many of the factors that child welfare workers take into consideration, such as the physical vulnerability of the child, the psychological functioning of the child or parent, or by social stress. This suggests that the prediction of injury using these simple factors as an intervention criterion may have questionable validity.