Almost 5,000 American pedestrians are killed annually, and 207,000 others injured. About one-fifth of injured pedestrians are children. One major reason children have increased pedestrian injury risk compared to adults is because crossing a street requires sophisticated cognitive and perceptual processing, skills that develop during childhood. Previous research suggests children can learn to be safer pedestrians. In particular, interventions offering repeated practice at crossing a street hold promise to teach children the complex cognitive-perceptual skills required. Such training was traditionally delivered by adults at streetside locations, but recently scholars have explored the use of virtual reality (VR). VR features several advantages for pedestrian safety training: a safe computer-generated environment with realistic images and sounds, a feeling of immersion without risk of actual injury, systematic delivery and control of stimuli to customize training to individual skill, and an engaging and fun learning environment. Further, VR training can be administered with minimal adult supervision and monitoring. The present study extended previous research using VR to teach children pedestrian safety in two primary ways. First, previously-tested VR environments were not mobile. A system that is fixed to a given location curtails broad dissemination of the intervention. By contrast, a mobile system can be transferred to different schools or community centers where children receive intense training over the course of a few weeks. We therefore refined a previously-developed and validated system to a more mobile VR environment. Second, we sought to conduct a pragmatic trial by testing VR training in a field environment under the usual circumstances of its potential implementation. Rather than training children in a sterile and artificial laboratory environment, we trained children in their schools and community centers, with the inevitable circumstances that arise in community environments. Using a within subjects pre-post research design, we evaluated children’s pedestrian safety at baseline, exposed them to six 15-minute VR training sessions, and re-evaluated pedestrian safety. We hypothesized pedestrian performance would improve across four performance measures: attention to traffic, delay in entering safe traffic gaps, time to contact with oncoming traffic while crossing, and unsafe simulated crossings. As hypothesized, children’s delays to enter traffic decreased slightly and significantly, implying the possibility of more rapid decision-making about gap safety following training. Children’s attention to traffic and time to contact by oncoming vehicles both decreased slightly but significantly following training, implying the possibility of slightly greater risk of pedestrian injury following training. There were no significant changes in the rate of unsafe crossings following training. A possible explanation for the pattern of results is that post-training, children made crossing decisions more confidently and more efficiently without sacrificing safety. They may have chosen equally safe but tighter gaps rather than waiting for obviously safe gaps. Posthoc analyses support this possibility. This study supports use of VR to teach child pedestrian safety and extends previous findings by installing a semi-mobile VR reality into community settings and implementing a pragmatic trial to test learning. The trial, which used six 15-minute training sessions, seemed to improve children’s efficiency in pedestrian crossings but not safety. More intense training, and more replication of the cognitive-perceptual process of crossing streets and receiving feedback about crossing safety, may be required to train children fully. Future research should consider dosage affects — how much training is required to teach children pedestrian skill? Development of VR systems that offer cost and portability for broad dissemination also should be prioritized. (Author/publisher)
Samenvatting