The Risk of a Safety-Critical Event Associated With Mobile Device Use in Specific Driving Contexts
Traffic Injury Prevention
The NSTSCE stakeholders have jointly funded this research for the purpose of developing and disseminating advanced transportation safety techniques and innovations. i EXECUTIVE SUMMARY Using a cell phone while driving has been associated with an increased crash risk. However, when examining the inherent subtasks, naturalistic driving studies (NDSs) have shown that safety-critical event (SCE) risk is associated with visual-manual subtaskswhich take the driver's eyes off the roadwayand not the
... versation itself. This study consisted of an analysis of data from NDSs involving commercial motor vehicle (CMV) drivers and light vehicle (LV) drivers. The first data set used was analyzed in Olson et al.'s 2009 case-control study of CMV driver distraction. (10) This data set comprised 4,451 SCEs (i.e., crashes, near-crashes, crash-relevant conflicts, and unintentional lane deviations) as well as 19,888 baseline epochs that were randomly sampled based on the time driven by each driver. The second data set was the 100-Car Study data set analyzed in Klauer et al.'s 2006 case-control study of LV driver distraction. (9) This data set comprised 828 SCEs (i.e., crashes and nearcrashes). The baseline epochs used, however, were the 17,213 epochs prepared in Guo & Hankey (28) because they were randomly sampled based on the time driven by each driver, as opposed to driver representation in the SCE sample. (9) The NDS data sets were partitioned into subsets representative of specific driving contexts. Groupings of (1) level of service, an ordinal measure of traffic density, (29) (2) relation to intersections and merge ramps, and (3) combinations of these two factors were prepared. The data were then "flagged" whenever specific mobile device subtasks were observed. Mobile device usage was investigated by summing all SCEs and baselines for which a subtask was observed in each context and comparing the likelihood of the subtask occurring in each context using chi-squared tests. Odds ratios and their respective 95% confidence intervals were then computed for mobile device subtasks in each context to investigate the association between their presence and the occurrence of an SCE. It was found that CMV and LV drivers varied as to how much they conversed on a mobile device, but did not vary their engagement in visual-manual subtasks, across the driving contexts examined. Furthermore, CMV drivers conversed less frequently when driving task demands were great, and LV drivers did not. The risk of an SCE associated with mobile device use (collapsed across subtasks) was dependent on the driving context as well as each subtask's associated SCE risk. Only visual-manual subtasks were associated with an increased SCE risk, while conversing was associated with a decreased risk. Overall, the study shows that drivers' engagement in mobile device subtasks, and the associated SCE risk, varies by driving context. The findings can be used to inform the design of in-vehicle interfaces that mitigate distraction by preventing visual-manual subtasks while driving.