Research into the relationship between pedestrian road accidents and the ability to predict vehicular motion is complicated by the fact that individuals may become accident victims for many reasons other than their ability to predict motion. The paper reports an attempt to solve the resultant methodological problem for selecting a sample of victims. The records of 276 child pedestrian accidents resulting in slight injuries in Wakefield and Leeds in 1972 were scrutinised and the following information obtained: (1) 74 per cent of accidents happened within 1/2 a mile of home; (2) 61 per cent of accidents happened when the victim was 'unmasked' by stationary vehicles; (3) very few accidents happened under adverse lighting and weather conditions (5 per cent and 18 per cent respectively); (4) 60 per cent of accidents were characterised by the possibility of inattention on the victim's part; (5) over 50 per cent of children were unaccompanied at the time of the accident; (6) in 70 per cent of accidents the child was held responsible for the accident; (7) highest accident frequency was found in the 6 and 7 year age groups; (8) there was no overall sex difference in accident rate over the age range 5 - 15; (9) although some districts of the Leeds-Wakefield area showed a high accident frequency there was no sex difference with respect to these districts. The above information was used to classify each accident in terms of the likelihood that prediction of motion was the major causal factor involved. For half of these accidents it was at least possible that poor motion prediction was largely responsible for the accidents occurring. To obtain a sample of 'victims with motion prediction difficulties', a victims group of approximately twice the size of the final sample would therefore have to be collected. Accident frequencies in Wakefield and Leeds indicate that such sampling would be feasible over a 1 - 2 year period. (Author/publisher)
Abstract