Development of Research Problem Statements That Utilize Naturalistic Driving Data to Improve Teen Driving Safety (2022)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

3 C H A P T E R 1 Background A persistent traffic safety priority is developing countermeasures to reduce young/novice driver crash involvement. In 2016, 15- to 20-year-olds represented 9% of all drivers involved in fatal crashes, but only 5.4% of the licensed drivers in the U.S. (NCSA 2018). Most prominent and successful among strategies to address this problem has been the application and evolution of graduated driver licensing (GDL) programs which are designed to increase beginning drivers’ experience and to limit their exposure to high-risk settings, primarily nighttime driving and the presence of other teenage occupants. Specific restrictions vary considerably from State to State. Compliance with GDL restrictions is key to reducing risky behaviors in a teen’s early driving experience; in many States restrictions are not lifted until age 18 (IIHS 2018). Fatal crash data analyses show that, relative to adult drivers, teen drivers have higher crash rates, are most likely to be involved in single-vehicle crashes, and are more likely to be speeding at the time of the crash (Bingham and Ehsani 2012; Gonzales et al. 2005; Shults et al. 2017). Researchers have concluded that many teen crashes occur as a result of inattention, distraction, and driving too fast for conditions (Carney et al. 2015, 2016; Curry et al. 2011; Braitman et al. 2008). Naturalistic driving data extends our understanding of (risky) teen driving behavior beyond what can be gleaned from crash analyses, in a number of important ways. It offers the potential to examine behavior on a continuous basis, instead of being limited to an event-based ‘snapshot’; and, critically, it provides an objective record of behavior in contrast to the subjective reports of drivers or after-the-fact inferences about risky behavior that been entered on a police report. Further, naturalistic driving data permits the analysis of (teen driver) behavior in relation to other road users and in the context of a particular set of operating (i.e., traffic, roadway, environmental, in-vehicle) conditions. Most prominent among naturalistic driving data collection efforts in this area is the second Strategic Highway Research Program Naturalistic Driving Study (SHRP2 NDS). The SHRP2 NDS includes 553 drivers aged 16-19, whose personal vehicles were instrumented with a data acquisition system that included accelerometers, cameras, GPS, forward radar, and vehicle network connections. This data source contains records of 410 crashes and 1182 near crashes by drivers in this age group, according to an Insight query on 09/25/2020. The SHRP2 NDS also includes at least one randomly selected 21-second ‘baseline’ trip segment for each driver where no crash, near crash, crash-relevant, or non-subject conflict event occurred. This is significant because it is only for trips in which an event (e.g., crash) occurred, or which are selected as baseline events, that video coders at the institution serving as custodian for the database analyze the event using the in-vehicle camera. This information is crucial for analyses of behaviors that place teens at elevated crash risk, e.g., distracted driving. For comparison, the total number of trips by drivers age 16-19 in the SHRP2 NDS database is 763,494; but analyses of driver behavior for these trips must rely on the (extensive) set of kinematic measures from the sensors noted above, plus camera views (though some camera angles such as those capturing images from inside the vehicle require viewing at a secure facility). These data, too, can be meaningful. A great deal of the published research on teen driving behavior—assessing compliance with GDL; examining

4 factors associated with distraction; crash analysis, including (exposure-based) crash rates; evaluating countermeasure effectiveness—has employed subjective data as dependent measures: questionnaire and survey responses, self-reports/driving diaries, etc. The SHRP2 NDS data describing—for example— maximum and mean speed, longitudinal and lateral acceleration rates, brake activations, maximum/minimum/mean distance to a lead vehicle, and percent of driving time/trip with headways < 3.5 seconds offer an unprecedented opportunity to operationalize risky driving through continuous, objective, real-time measures. The objective in this Phase 1 research project was to determine if and how the NDS database can be exploited to support an agenda for teen driver countermeasure development and evaluation. With an affirmative assessment of its potential in this regard, the scope of work described below was designed to produce a roadmap for next steps in the form of concrete and viable options for continuing research in this critical traffic safety area. The scope of work in this project included a series of tasks in which our research team first examined the NDS data attributes and their strengths and weaknesses for teen driver behavioral analyses. Next, we prepared an inventory of data elements deemed most helpful for advancing teen driver research, also considering supplementary data sources that could address critical gaps in the SHRP2 NDS. We relied on the findings of a recent literature review, and responses by teen driving safety experts to a survey administered during this study, to establish a set of priority research questions; these were grounded in the framework of graduated driver licensing (GDL) programs and theory. The final project task was directed to the highest priority research questions for which the richest array of elements within the NDS database could be identified, culminating in three RNSs prepared according to TRB guidelines, each complemented with a full Data Specification as required to support an extraction and processing request under a Data Use License (DUL) application to the SHRP2 NDS data custodian.