Learning to drive: A reconceptualization

Promoting transportation safety in adolescence: the drivingly randomized controlled trial

BACKGROUND

The impact of young drivers' motor vehicle crashes (MVC) is substantial, with young drivers constituting only 14% of the US population, but contributing to 30% of all fatal and nonfatal injuries due to MVCs and 35% ($25 billion) of the all medical and lost productivity costs. The current best-practice policy approach, Graduated Driver Licensing (GDL) programs, are effective primarily by delaying licensure and restricting crash opportunity. There is a critical need for interventions that target families to complement GDL. Consequently, we will determine if a comprehensive parent-teen intervention, the Drivingly Program, reduces teens' risk for a police-reported MVC in the first 12 months of licensure. Drivingly is based on strong preliminary data and targets multiple risk and protective factors by delivering intervention content to teens, and their parents, at the learner and early independent licensing phases.

METHODS

Eligible participants are aged 16-17.33 years of age, have a learner's permit in Pennsylvania, have practiced no more than 10 h, and have at least one parent/caregiver supervising. Participants are recruited from the general community and through the Children's Hospital of Philadelphia's Recruitment Enhancement Core. Teen-parent dyads are randomized 1:1 to Drivingly or usual practice control group. Drivingly participants receive access to an online curriculum which has 16 lessons for parents and 13 for teens and an online logbook; website usage is tracked. Parents receive two, brief, psychoeducational sessions with a trained health coach and teens receive an on-road driving intervention and feedback session after 4.5 months in the study and access to DriverZed, the AAA Foundation's online hazard training program. Teens complete surveys at baseline, 3 months post-baseline, at licensure, 3months post-licensure, 6 months post-licensure, and 12 months post-licensure. Parents complete surveys at baseline, 3 months post-baseline, and at teen licensure. The primary end-point is police-reported MVCs within the first 12 months of licensure; crash data are provided by the Pennsylvania Department of Transportation.

DISCUSSION

Most evaluations of teen driver safety programs have significant methodological limitations including lack of random assignment, insufficient statistical power, and reliance on self-reported MVCs instead of police reports. Results will identify pragmatic and sustainable solutions for MVC prevention in adolescence.

TRIAL REGISTRATION

ClinicalTrials.gov # NCT03639753.

Tools for Transport: Driven to Learn With Connected Vehicles

Vehicle automation and assistance technologies have been touted as a means to reduce traffic collisions by minimizing or eliminating "error-prone" and inefficient human operators. In concept, automation exists on a continuum that includes engaged driving by a human operator augmented by automated support features, vigilant driver monitoring of vehicle behavior with the possibility of driver take-over, to full automation with no active monitoring by a human operator. Moreover, the degree of automation varies by vehicle features (e.g., lane centering, emergency braking, adaptive cruise control, parking), by setting, meaning that automated features may or may not be available depending on specific attributes of the traffic environment (e.g., traffic volume, road geometry, etc), and by implementation (e.g., haptic vs. auditory warnings). Thus, these automotive "transportation tools" are highly heterogeneous and pose unique challenges and opportunities for driver training. In this paper, we report the results of an experimental study (n = 36) to determine if enhanced vehicle feedback influences driver trust, effort, frustration, and performance (indexed by reaction time) in a virtual driving environment. Results are contextualized in the extant literature on learning to operate motor vehicles and outline key research questions essential for understanding the processes by which skilled performance develops with respect to a real-world practical tool: the increasingly automated automobile.

Advancing our understanding of cognitive development and motor vehicle crash risk: A multiverse representation analysis

Neurobiological and cognitive maturational models are the dominant theoretical account of adolescents' risk-taking behavior. Both the protracted development of working memory (WM) through adolescence, as well as individual differences in WM capacity have been theorized to be related to risk-taking behavior, including reckless driving. In a cohort study of 84 adolescent drivers Walshe et al. (2019) found adolescents who crashed had an attenuated trajectory of WM growth compared to adolescent drivers who never reported being in a crash, but observed no difference in WM capacity at baseline. The objectives of this report were to attempt to replicate these associations and to evaluate their robustness using a hybrid multiverse - specification curve analysis approach, henceforth called multiverse representation analysis (MRA). The authors of the original report provided their data: 84 adolescent drivers with annual evaluations of WM and other risk factors from 2005 to 2013, and of driving experiences in 2015. The original analysis was implemented as described in the original report. An MRA approach was used to evaluate the robustness of the association between developmental trajectories of WM and adolescents' risk-taking (indexed by motor vehicle crash involvement) to different reasonable methodological choices. We enumerated 6 reasonable choice points in data processing-analysis configurations: (1) model type: latent growth or multi-level regression, (2) treatment of WM data; (3) which waves are included; (4) covariate treatment; (5) how time is coded; and (6) link function/estimation method: weighted least squares means and variance estimation (WLSMV) with a linear link versus logistic regression with maximum likelihood estimation. This multiverse consists of 96 latent growth models and 18 multi-level regression models.