Abstract
Data from police accident reports involving pedestrians less than 20 years of age in Hartford, Conn, during 1988 through 1990 were abstracted and entered into a geographic information system. Two high-frequency collision areas were identified and compared. There were 374 child pedestrians involved in collisions (a rate of 28 per 10,000). Two high-occurrence areas accounted for 30% of collisions. Collisions in one of these areas were more likely to involve younger children (8.1 vs 10.2 years of age) and to occur in the late afternoon, and occurred closer to the child's residence, than collisions in the other area. The geographic information system is a useful tool in the study of child pedestrian collisions.