Young drivers' perceptions of risk and difficulty: Day versus night.

Inexperienced drivers have a higher crash risk than others, particularly at night when drivers of all ages are at increased risk, but there has been little if any research on day-night differences in drivers' perceptions of risk and difficulty. Also, previous research on determinants of risk and difficulty ratings has focused largely on effects of vehicle speed, and researchers have identified a need to look more broadly at how ratings are affected by the complexity of depicted road-traffic situations. This study addressed those issues, using an exploratory approach entailing both quantitative and qualitative analyses. The 62 participants (28 females, 34 males) were grouped according to level of driving experience and age (16-30 years). Determinants of ratings were investigated using 6 day-night pairs of video clips at 6 different locations, plus 2 night-only clips at 2 other locations (14 clips in all). Depicted driving situations varied naturalistically in complexity as well as speed. They were presented in random orders twice to each participant, who used a slider to rate each clip continuously, once for risk and once for difficulty in balanced order. Consistent with previous evidence, there were no significant differences in ratings between participant groups. Ratings of risk and difficulty were highly correlated, as expected, but Rasch analysis confirmed that they reflected empirically different constructs. Ratings were significantly higher for night than day. They also differed significantly between locations, with vehicle speed not being their main determinant. Further, there were significant differences between locations in the relationship between risk and difficulty ratings, suggesting that participants' interpretations of 'risk' and 'difficulty' varied between locations. Qualitative analysis of rating patterns in relation to video clip content utilised constructs drawn from information theory and research on human information processing. It was concluded that more systematic research is required to investigate effects on ratings of variables determining the amount of information drivers must process, such as numbers and types of other road users and the predictability of their actions, road types and configurations, visibility conditions, particularly day versus night - as well as effects of vehicle speed which determines the required rate of processing.