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.

14 C H A P T E R 3 Research Findings and Applications The findings in this project consist of Research Needs Statements (RNSs) prepared by our team for the questions to which our focus was narrowed as described above. Each RNS was prepared according to guidance found at the TRB RNS website http://rns.trb.org. Augmenting each RNS is a Data Specification consistent with what is required in a DUL application to the data custodian. This identifies every SHRP2 NDS variable and data element, the table from which it will be extracted, and any processing necessary to derive information that cannot be read directly from a table. If a need to view any PII in a secure data enclave is anticipated, this is noted explicitly. While developing the Data Specification for each RNS, the nature and availability of information in the NDS suggested slight modifications to the four research questions prioritized above, typically to phrase the question with greater specificity; these changes are reflected in the proposed RNS titles. In addition, during this effort it became apparent that there was a very substantial overlap in the data extractions related to the two of these questions – addressing, respectively, the extent to which crash contributing factors such as distracted driving are involved in teen crashes, and how this changes with experience – such that these could effectively be addressed via analyses under a single RNS. Given the emphasis on GDL as the framework within which the recommended research agenda was developed, it also deserves mention that investigations addressing teens’ compliance with GDL restrictions, and how risk changes for those who are noncompliant, are effectively ruled out by the strict privacy protections afforded NDS participants. Specifically, a driver’s GDL status during study enrollment is not known. Communication with the data custodian has revealed that this information can only be made available with signed consent from each participant of interest. If researchers were to obtain both permission from participants and license information from the corresponding State DOT (including the date the GDL restrictions were lifted, should that time period fall during study enrollment), the most accurate compliance information could be derived for nighttime driving restrictions (for any trip via the time series data or Trip Summary table using trip start and end time). The presence of passengers is only known for trip segments in the event table, which include baseline (non-safety critical) events, but passenger age is not known and would have to be estimated through video review at a secure data viewing facility. These RNSs follow: 1. How and to what extent is teens’ exposure to diversity of traffic/road environments in their early driving experience related to driver performance, including but not limited to crash/near crash involvement? 2. To what extent do driving risks increase (or differ) for vulnerable teen populations, specifically those with an indication of ADHD, apart from the influence of other driver factors? 3. How and to what extent does driving distraction (internal and external) contribute to teens’ driving performance, including crashes and near crashes, and do these relationships change as teens gain driving experience?

15 Research Needs Statement 1: How and to what extent is teens’ exposure to diversity of traffic/road environments in their early driving experience related to driver performance, including but not limited to crash/near crash involvement? RESEARCH NEEDS STATEMENT How and to what extent is teens’ exposure to diversity of traffic/road environments in their early driving experience related to driver performance, including but not limited to crash/near crash involvement? Problem An important question is whether teenagers who are exposed to greater diversity of traffic and road environments early in their driving career have lower crash involvement than those who are exposed to less diversity. Ideally, this question would be addressed by analyzing how driving exposure—both the amount of driving and driving conditions—change when teens make the transition from supervised to unsupervised driving. While supervised driving data are not available in NDS, it is possible to compare the diversity of experience in the early months of unsupervised driving with later months, and to examine the association of exposure to greater diversity with crashes and near crashes. The period of early exposure during unsupervised driving will be the focus of the proposed research. A research hypothesis will be that teens who spend a greater amount of either time or miles driving in a wider range of environments in their first 6 months of unsupervised driving are significantly less likely to experience a crash or near crash in each subsequent 6-month experience band, compared to teens with a more restricted range of experience in their first 6 months of unsupervised driving. If it is found that not enough participants were enrolled in the NDS during their first 6 months of driving experience, this hypothesis may be modified to use the first 12 months of driving experience as the reference interval. Objective The objective is to evaluate how exposure to greater diversity in traffic and road environments is associated with driver performance indicators such as crashes or near crashes. Diversity of driving environments can be operationalized using variables from the NDS Trip Summary table, which will be collapsed by data custodians across experience bands as cumulative total, minimum, maximum, and average for each driver. At a minimum these variables are to include time and distance with varying numbers of radar targets, to serve as an indication of traffic density; time and distance spent driving at various speeds; the proportion of each trip spent on roadways with varying speed limits; and the proportion of each trip spent driving on varying roadway classifications (e.g., rural freeway, urban two lane, etc.). Daytime versus nighttime driving, classified according to the (local) time of day that the trip started, is also to be considered.

16 Key Words Novice, teen, driver, experience, exposure, road, traffic, safety, crash, near crash, event Related Work Related work has focused on teens’ supervised driving experience. There is evidence that supervised practice driving is characterized mainly by routine trips in relatively benign, undemanding conditions, as opposed to potentially challenging situations such as highways, darkness, heavy traffic, or country roads.1,2 Urgency/Priority High Cost / Period of Performance $85,000 (assuming ≤ $10,000 for VTTI fees) / 12 months User Community Driver educators, GDL program officials, caregivers, other researchers Implementation These research findings should be distributed through TRB (TR News, TRR Journal, and/or an Annual Meeting presentation), other peer-review journals, other conferences (e.g., Lifesavers, ADTSEA), and avenues for communicating with States available to GHSA. Effectiveness Expected outcomes include a better understanding of how more diverse early driving experience lowers risks for teens, with a translation to education and regulation as measured through updates to materials used for training curricula and revised GDL requirements. Also expected is a greater appreciation in the traffic safety community of the roles naturalistic driving data collection can play both in developing safe driving habits and in evaluating driver capabilities and performance, potentially stimulating the use of increasingly affordable technologies for capturing such data by trainers/educators, caregivers, and others. References 1. Goodwin, A.H., L.H. Margolis, and M. Waller. 2010. Parents, Teens and the Learner Stage of Graduated Licensing. AAA Foundation for Traffic Safety, Washington, DC. https://aaafoundation.org/parents-teens-learner-stage-graduated-driver-licensing/ (As of April 30, 2020). 2. Ehsani, J. P., S.G. Klauer, C. Zhu, P. Gershon, T.A. Dingus, and B.G. Simons-Morton. 2017. “Naturalistic Assessment of the Learner License Period.” Accident Analysis & Prevention, Vol. 106, pp. 275-284.

17 Data Specification Notes: The ‘Participant Receive License’ variable is a discrete measure, so drivers with the values “18+” and “<15” will be excluded from this analysis; since the numeric age of licensure is not known in the NDS, it will not be possible to calculate an accurate experience band for those drivers. Data filter: Driver age < 20 and ‘Participant Receive License’ not equal to 18+ or < 15 Variables to request under Data Use License (italicized variables are derived by VTTI) SHRP2 Data Selection SHRP2 Variable VTTI Calculation Driver Experience Band Participant Receive License minus Driver age at trip to nearest half year (rounded down, consistent with FARS, calculated by VTTI using driver date of birth) rounded to the nearest half year to create 6-month experience bands. All trips are sorted according to the 6-month experience band each falls into. Trip Summary Trip Count Total number of trips per experience band Trip Summary time where radar targets = 1 Cumulative, min, max, avg for each experience band for each driver time where radar targets = 2 Cumulative, min, max, avg for each experience band for each driver time where radar targets = 3 Cumulative, min, max, avg for each experience band for each driver time where radar targets = 4 Cumulative, min, max, avg for each experience band for each driver time where radar targets = 5 Cumulative, min, max, avg for each experience band for each driver time where radar targets = 6+ Cumulative, min, max, avg for each experience band for each driver distance where radar targets = 1 Cumulative, min, max, avg for each experience band for each driver distance where radar targets = 2 Cumulative, min, max, avg for each experience band for each driver distance where radar targets = 3 Cumulative, min, max, avg for each experience band for each driver distance where radar targets = 4 Cumulative, min, max, avg for each experience band for each driver distance where radar targets = 5 Cumulative, min, max, avg for each experience band for each driver distance where radar targets = 6+ Cumulative, min, max, avg for each experience band for each driver time at 0-10 mph Cumulative, min, max, avg for each experience band for each driver time at 10-20 mph Cumulative, min, max, avg for each experience band for each driver time at 20-30 mph Cumulative, min, max, avg for each experience band for each driver

18 SHRP2 Data Selection SHRP2 Variable VTTI Calculation time at 30-40 mph Cumulative, min, max, avg for each experience band for each driver time at 40-50 mph Cumulative, min, max, avg for each experience band for each driver time at 50-60 mph Cumulative, min, max, avg for each experience band for each driver time at 60-70 mph Cumulative, min, max, avg for each experience band for each driver time at 70-80 mph Cumulative, min, max, avg for each experience band for each driver time at >80 mph Cumulative, min, max, avg for each experience band for each driver distance at 0-10 mph Cumulative, min, max, avg for each experience band for each driver distance at 10-20 mph Cumulative, min, max, avg for each experience band for each driver distance at 20-30 mph Cumulative, min, max, avg for each experience band for each driver distance at 30-40 mph Cumulative, min, max, avg for each experience band for each driver distance at 40-50 mph Cumulative, min, max, avg for each experience band for each driver distance at 50-60 mph Cumulative, min, max, avg for each experience band for each driver distance at 60-70 mph Cumulative, min, max, avg for each experience band for each driver distance at 70-80 mph Cumulative, min, max, avg for each experience band for each driver distance at >80 mph Cumulative, min, max, avg for each experience band for each driver % Urb Frwy Cumulative, min, max, avg for each experience band for each driver % Urb Frwy < 4 Lns Cumulative, min, max, avg for each experience band for each driver % Urb 2 Ln Cumulative, min, max, avg for each experience band for each driver % Urb Multi Div Non-Frwy Cumulative, min, max, avg for each experience band for each driver % Urb Multi Undiv Non-Frwy Cumulative, min, max, avg for each experience band for each driver % Rur Frwy Cumulative, min, max, avg for each experience band for each driver % Rur Frwy < 4 Lns Cumulative, min, max, avg for each experience band for each driver % Rur 2 Ln Cumulative, min, max, avg for each experience band for each driver % Rur Multi Div Non-Frwy Cumulative, min, max, avg for each experience band for each driver % Rur Multi Undiv Non-Frwy Cumulative, min, max, avg for each experience band for each driver % Other Class Cumulative, min, max, avg for each experience band for each driver

19 SHRP2 Data Selection SHRP2 Variable VTTI Calculation % Spd Lim 35 or Less Cumulative, min, max, avg for each experience band for each driver % Spd Lim 40-50 Cumulative, min, max, avg for each experience band for each driver % Spd Lim 55-65 Cumulative, min, max, avg for each experience band for each driver % Spd Lim 70 or Greater Cumulative, min, max, avg for each experience band for each driver Day or Night Day - Trip start local time hour of day = 5:00 am to 8:59 pm Night - Trip start local time hour of day = 9 p.m. to 4:59 am Event detail Event Severity 1 Crash y/n y = driver was involved in at least one crash during the corresponding experience band Near crash y/n y = driver was involved in at least one near crash during the corresponding experience band

20 Research Needs Statement 2: To what extent do driving risks increase (or differ) for vulnerable teen populations, specifically those with an indication of ADHD, apart from the influence of other driver factors? RESEARCH NEEDS STATEMENT To what extent do driving risks increase (or differ) for vulnerable teen populations, specifically those with an indication of ADHD, apart from the influence of other driver factors? Problem Young drivers with developmental disabilities may be at increased risk for crashes due to impairments commonly associated these conditions. In recent years, a growing body of research has examined driving risks for teens on the autism spectrum, including those with attention deficit hyperactivity disorder (ADHD). Specifically, research suggests teenagers with ADHD display a number of decrements in driving skill and are more likely to engage in distracted driving and to drive at higher speeds.1 Objective The objective is to compare the exposure-based crash and near crash involvement rates as well as self-reported risky driving behaviors for teen drivers with higher versus lower ADHD screen scores, taking into account the potential influence of other behavioral and demographic factors captured in NDS data. NDS participants completed Barkley’s ADHD Screening Test, which operationalizes ADHD symptoms in terms of specific behaviors, during study enrollment. The sum across all six items on this questionnaire indicates whether an individual may have ADHD (a sum of 7 or greater indicates possible ADHD). To meet the present objective, teen participants may be grouped dichotomously, contrasted according to highest versus lowest quartiles, or otherwise as researchers deem appropriate. Risky driving behaviors may be gauged via the Risk Taking Questionnaire; sleep habits, caffeine intake, tobacco and alcohol use, occupation, work habits, and presence of children in the home via the Sleep Habits Questionnaire; and sensory stimulation preferences via the Sensation Seeking Scale Survey. Additional factors to be considered include, but are not limited to, prescribed medications, and demographic factors such as age, gender, education, race, ethnicity, household status (two- or one-parent or live alone), household income, and work status. Key Words Novice, teen, driver, safety, crash, near crash, event, autism spectrum, ADHD Related Work Related work has focused more broadly on teens with autism spectrum disorder (ASD) not limited to ADHD. Rates of licensure are substantially lower among teenagers with ASD than teenagers without ASD.2 Although all novice drivers are at increased risk because of inexperience, there is evidence that teenagers with ASD face additional challenges that include

21 difficulties identifying social hazards (e.g., pedestrians), slower reaction times, and poorer situation awareness skills.3 Urgency/Priority High Cost / Period of Performance $85,000 (assuming ≤ $10,000 for VTTI fees) / 12 months User Community Caregivers, occupational therapists, other researchers, licensing officials, law enforcement Implementation These research findings should be distributed through TRB (TR News, TRR Journal, and/or an Annual Meeting presentation), other peer-review journals, other conferences (e.g., ADED), and avenues for communicating with States available to GHSA. Effectiveness Expected outcomes include a better understanding of the degree and nature of additional risks experienced by this vulnerable group of teen drivers. This knowledge should not only inform these teens’ caregivers and those who provide training to them, but could benefit law enforcement personnel who encounter teen drivers with ADHD. Also expected is a greater appreciation in the traffic safety community of the roles naturalistic driving data collection can play in developing safe driving habits and in measuring driver capabilities and performance. References 1. Bishop, H., L. Boe, D. Stavrinos, and J. Mirman. 2018. “Driving Among Adolescents with Autism Spectrum Disorder and Attention-Deficit Hyperactivity Disorder.” Safety, Vol. 4, No. 3, p. 40. 2. Curry, A.E., B.E. Yerys, P. Huang, and K.B. Metzger. 2018. “Longitudinal Study of Driver Licensing Rates Among Adolescents and Young Adults with Autism Spectrum Disorder.” Autism, Vol. 22, No 4, pp. 479-488. 3. Silvi, C., B. Scott-Parker, and C. Jones. 2018. “A Literature Review of the Likely Effects of Autism Spectrum Disorder on Adolescent Driving Abilities.” Adolescent Research Review, Vol. 3, No. 4, pp. 449-465.

22 Data Specification Data filter: Driver age < 20 Variables to request under Data Use License (italicized variables are derived by VTTI) SHRP2 Data Selection SHRP2 Variable VTTI Calculation Driver - Barkley’s ADHD Screening Test Barkley’s Score Event detail Crash y/n y = driver was involved in at least one crash n = driver had no crashes during study enrollment Near crash y/n y = driver was involved in at least one near crash n = driver had no near crashes during study enrollment Crash Total Count of crashes per driver Near Crash Total Count of near crashes per driver Contributing Crash Total Count of crashes per driver where fault = ‘subject driver’ Contributing Near Crash Total Count of near crashes per driver where fault = ‘subject driver’ Driver Age at enrollment Education Gender Race Ethnicity Head of household [household] Income Number in household Work status Driver mileage last year Driver - Risk Taking Questionnaire ~ all variables Driver – Sensation Seeking Scale Survey SSS Total Score Disinhibition Summary Metric Thrill Seeking Summary Metric Boredom Summary Metric Experience Seeking Summary Metric Driver – Sleep Habits Questionnaire ~ all variables Driver – Medical Conditions & Medications Prescribed Medications

23 SHRP2 Data Selection SHRP2 Variable VTTI Calculation Trip Summary Total Miles Sum of each Trip Distance

24 Research Needs Statement 3: How and to what extent does driving distraction (internal and external) contribute to teens’ driving performance, including crashes and near crashes, and do these relationships change as teens gain driving experience? RESEARCH NEEDS STATEMENT How and to what extent does driving distraction (internal and external) contribute to teens’ driving performance, including crashes and near crashes, and do these relationships change as teens gain driving experience? Problem Distracted driving has become a growing concern over the past few decades with the advent of smartphones and other technologies that have the potential to divert attention from the task of driving. However, the contribution of distracted driving to crashes is not well established. It can be challenging for an officer to determine whether a driver was distracted at the time of crash. For that reason, it is widely believed that distractions are underreported in crash records. Given the limitations of crash data, researchers have turned to observational methods to examine the prevalence and increased risk posed by non-driving-related tasks. Naturalistic studies, most notably the SHRP2 NDS, can objectively identify driver distraction behavior immediately before a crash or other event. The SHRP2 NDS data provide an opportunity to address a number of questions related to teenagers and distracted driving: Which potentially distracting driver behaviors are most common among teenage drivers? Under what conditions do distracted driving behaviors most commonly occur (e.g., time of day, day of week, amount of traffic, presence of passengers)? Which distracted driver behaviors are most likely to contribute to crashes and near crashes? How does driver behavior change in presence of (teenage) passengers? These naturalistic data also support analyses of how distracting behaviors by teen drivers – and their role in crash causation – change as these novices gain experience behind the wheel. Objective The objective is to gauge the association between confirmed incidences of distracting behaviors and inattention to the driving task by teen drivers with crash and near crash involvement, in relation to their incidence during baseline events; and to determine whether these incidences contribute to crashes and near crashes, and if and how these relationships change with increasing driving experience. Such confirmation is to be provided via VTTI video coder entries for each trip segment in an event – whether it is a crash, near crash, or baseline event – that includes up to three driver behaviors (including distraction) and up to three secondary tasks that identify what the driver was distracted by (e.g., cell phone use, passenger interaction, external distractions). Researchers should pay particular attention to the secondary task outcome, signifying the coder’s determination of whether the secondary task contributed to the event sequence or severity.

25 Key Words Novice, teen, driver, safety, crash, near crash, event, distraction Related Work There is evidence that the risk of a crash or near crash for newly licensed teenage drivers increases substantially when the driver is dialing a cell phone, reaching for a cell phone, or sending or receiving text messages.1 Another (naturalistic) study of high school-age drivers found that loud conversation and horseplay among passengers was more strongly associated with near collisions and high g-force events than electronic device use and other potentially distracting behaviors.2 Many other studies have demonstrated that cell phone use and texting in particular is associated with decrements in driving performance.3,4 At the same time, research has shown that teens’ crash rates decline sharply during the first year of independent driving,5 potentially related to a reduced engagement in distracting behaviors. Urgency/Priority High Cost / Period of Performance $85,000 (assuming ≤ $10,000 for VTTI fees) / 12 months User Community Driver educators, GDL program officials, caregivers, other researchers Implementation These research findings should be distributed through TRB (TR News, TRR Journal, and/or an Annual Meeting presentation), other peer-review journals, other conferences (e.g., Lifesavers, ADTSEA), and avenues for communicating with States available to GHSA. Effectiveness Expected outcomes include a reliable accounting of the extent to which inattention and distraction contribute to teen crashes and near crashes, and whether these risk factors change with increased early driving experience. This information could translate to improvements in novice driver education and regulation, potentially introducing or increasing the use of affordable technologies for naturalistic driving data collection during the unsupervised driving phase of GDL programs and, more broadly, underscoring the utility of this technology in developing safe driving habits. References 1. Klauer, S.G., F. Guo, B.G. Simons-Morton, M.C. Ouimet, S.E. Lee, and T.A. Dingus. 2014. “Distracted Driving and Risk of Road Crashes Among Novice and Experienced Drivers.” New England Journal of Medicine, Vol. 370, No. 1, pp. 54-59

26 2. Goodwin, A. H., R. Foss, S.S. Harrell, and N.P. O'Brien. 2012. Distracted Driving Among Newly Licensed Teen Drivers. AAA Foundation for Traffic Safety. Washington, DC. https://aaafoundation.org/distracted-driving-among-newly-licensed-teen-drivers/ (As of April 30, 2020). 3. McCartt, A. T., L.A. Hellinga, and K.A. Braitman. 2006. “Cell Phones and Driving: Review of Research.” Traffic Injury Prevention, Vol.7, No. 2, pp. 89-106. 4. Caird, J.K., C.R. Willness, P. Steel, and C. Scialfa. 2008. “A Meta-Analysis of the Effects of Cell Phones on Driver Performance.” Accident Analysis & Prevention, Vol. 40, No 4, pp. 1282- 1293 5. McCartt, A. T., V.I. Shabanova, and W.A. Leaf. 2003. “Driving Experience, Crashes and Traffic Citations of Teenage Beginning Drivers.” Accident Analysis & Prevention, Vol. 35, No. 3, pp. 311-320. Data Specification Notes: This Data Specification assumes use of the Event Detail table to investigate differences in distraction-related crash contributing factors among teen drivers at varying levels of experience, relying on the secondary task and secondary task outcome variables (including approximately 70 different elements/distractions) where each event can have up to 3 secondary tasks identified. This will support analyses at both the group (i.e., internal distraction, external distraction) level and at the individual (e.g., cell phone use) secondary task level. The extent of contribution will be operationalized as a rate calculated as the incidence of a given secondary task contributing to an event in relation to the incidence of that secondary task not contributing to the event or occurring during a baseline event. Events will be sorted by experience band to gauge the change in the rate for distracting elements individually and at the group level. Data filter: Driver age < 20; Event Severity 1 = crash, near crash, baseline, and additional baseline Variables to request under Data Use License (italicized variables are derived by VTTI) SHRP2 Data Selection SHRP2 Variable VTTI Calculation Driver Age at Event Driver age at enrollment to nearest half year Experience Band Participant Receive License minus Age at Time of Event rounded to the nearest half year to create 6-month experience bands. Each event is classified according to its calculated experience band. Event Detail Event Severity 1 Secondary Task 1 Secondary Task 2 Secondary Task 3 Secondary Task 1 Outcome Secondary Task 2 Outcome Secondary Task 3 Outcome