Innovations in using virtual reality to study how children cross streets in traffic: evidence for evasive action skills

Purpose Children in middle childhood are at an increased risk for injury in pedestrian environments. This study examined whether they are capable of showing evasive action (ie, adjusting crossing speed) to avoid injury when crossing streets. Methods The study used a fully immersive virtual reality (VR) system interfaced with a three-dimensional movement measurement system so that the actual crossing behaviour of 7–10-year-old children under different traffic conditions could be precisely measured. Relating outcomes to that which would have been obtained based on using the approach of estimating walking speed and assuming a constant speed provided insights into the realised benefits of the current movement monitoring VR system. Results Controlling for age and sex, children showed evasive action, crossing more quickly as traffic conditions became more risky. Using an average and assuming a constant walking speed underestimated actual walking speed, failing to capture evasive action and leading to overestimation of children being hit compared with the actual incidence of hits. Conclusions VR technology is a valuable tool for assessing child pedestrian behaviour. However, systems need to allow the child to cross the street so their level of pedestrian skill is appropriately measured. The current findings provide the first evidence that children are capable of implementing evasive action in reaction to risky traffic conditions.