Abstract
This study investigates the use of land use variables in predicting the number of child pedestrian accident casualties taking Newcastle upon Tyne (UK) as a case study. GIS techniques are used to create spatial models, from which generalised linear models (GLM) are developed over all child accidents and KSI (Killed or Seriously Injured) considering the child pedestrian casualty numbers, land use trip attractors and generators as variables. The results show that that secondary retail and high density residential were the main land use types associated with child pedestrian casualties, in such a way that the former was positively associated while the latter had a negative association. It also found that educational sites were also positively associated with child pedestrian casualties, especially for the KSI.
