Multilevel modeling of gap-selection behaviors: Effects of vehicle arrival time and personal walking time projection abilities in mixed traffic conditions

Previous studies have attributed the relatively high risk of elderly pedestrians to their declining abilities, particularly the ability to estimate their own walking time and time-to-arrival of oncoming vehicles. However, how these two abilities affect street-crossing safety in mixed traffic conditions remains unclear. This study conducted a simulation study using filmed real-life traffic to investigate this issue. Repeated measures of gap-selection decision and safety outcomes among 82 adults (42 elderly participants aged 60 years or older and 40 aged between 20 and 45 years) were collected. Two sets of scenarios were developed: one car–one lane simple-flow tasks and multiple vehicle–one lane complex-flow tasks. The analysis results demonstrated that in simple- and complex-flow tasks these two abilities are both critical to gap-selection safety; however, pedestrians’ reliance on walking-time projection is more consistent than is their reliance on time to arrival (TTA) projection when making gap-selection decisions. Although young pedestrians are safer on average than elderly pedestrians, how they respond to oncoming vehicles is similar; in particular, both young and elderly pedestrians tend to select longer gaps even when oncoming vehicles are moving quickly. The responses of the elderly pedestrians were relatively varied and skewed toward being unsafe, and thus their on-average gap-selection performance was inferior to that of the young pedestrians. The benefit of offering countdowns to oncoming vehicle arrivals, as a means of enhancing gap-selection safety, is appropriate when traffic flow is simple but relatively less so in complex-flow scenarios, indicating the importance of considering information overload in countermeasure design.


  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 01682970
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Oct 5 2018 3:06PM